JOURNAL OF BUSINESS AND ACCOUNTING
Transcript of JOURNAL OF BUSINESS AND ACCOUNTING
JOURNAL OF BUSINESS AND ACCOUNTING
Volume 12, Number 1 ISSN 2153-6252 FALL 2019
IN THIS ISSUE
Oil Reserves and Financial Information During 2011 - 2015
………………………………….Anwar Salimi
Does the United States Postal Service Charge Amazon Too Little?
……………………….Barton and MacArthur
Underfunded Pension Liabilities’ Impact on Security Returns: The U.S. Versus the European Union
………………………………………Ronald A. Stunda
Technostress: An Antecedent of Job Turnover Intention in the Accounting Profession
……………………………..Stacy Boyer-Davis
The Comprehensive Taxation System Existing During the Roman Empire
………………………………….King, Case and Roosa
Keeping the Oil and Gas Pipelines Operating as Assets: Safety Issues Can Make Them Liabilities
……………………………………….Carol Sullivan
An Analysis of Alternative Work Arrangements to Address the Gender Gap in Public Accounting
…………………………………Anderson and Smith
Variations in Unfunded Pension Liabilities Across U.S. States …………………………….Paul A. Tomolonis
Consistency of Ethical Analysis Among Accounting Students
…………………………………..Flynn, Deno and Buchan
Fraud-Detecting Effectiveness of Management and Employee Red Flags Aa Perceived by Three
Different Groups of Professionals
………………………………………Moyes, Anandarajan and Arnold
A REFEREED PUBLICATION OF THE AMERICAN SOCIETY
OF BUSINESS AND BEHAVIORAL SCIENCES
2
JOURNAL OF BUSINESS AND ACCOUNTING P.O. Box 502147, San Diego, CA 92150-2147: Tel 909-648-2120
Email: [email protected] http://www.asbbs.org
ISSN: 2153-6252 Editor-in-Chief
Wali I. Mondal
National University
Editorial Board
Karen Blotnicky
Mount Saint Vincent University
Pani Chakrapani
University of Redlands
Gerald Calvasina
University of Southern Utah
Shamsul Chowdhury
Roosevelt University
Steve Dunphy
Indiana University Northeast
Lisa Flynn
SUNY, Oneonta
Sharon Heilmann
Wright State University
Sheldon Smith
Utah Valley University
Mohammed Shaki
Saint Leo University
Saiful Huq
University of New Brunswick
William J. Kehoe
University of Virginia
Maureen Nixon
South University Virginia Beach
Z.S. Andrew Demirdjian
California State University long Beach
Douglas McCabe
Georgetown University
Marybeth McCabe
National University
Thomas Vogel
Canisius College
Jake Zhu
California State University San
Bernardino
Linda Whitten
Skyline College
The Journal of Business and Educational Leadership is a publication of the American
Society of Business and Behavioral Sciences (ASBBS). Papers published in the Journal
went through a blind review process prior to acceptance for publication. The editors wish
to thank anonymous referees for their contributions.
The national annual meeting of ASBBS is held in Las Vegas in March of each year and the
international meeting is held in June of each year. Visit www.asbbs.org for information
regarding ASBBS.
3
JOURNAL OF BUSINESS AND ACCOUNTING
ISSN 2153-6252
Volume 12, Number 1 Fall 2019
TABLE OF CONTENTS
Oil Reserves and Financial Information During 2011 - 2015
Anwar Salimi……………………………………….4
Does the United States Postal Service Charge Amazon Too Little?
Barton and MacArthur……………………………….19
Underfunded Pension Liabilities’ Impact on Security Returns: The U.S. Versus the
European Union
Ronald A. Stunda………………………………………36
Technostress: An Antecedent of Job Turnover Intention in the Accounting
Profession
Stacy Boyer-Davis…………………………………………49
The Comprehensive Taxation System Existing During the Roman Empire
King, Case and Roosa………………………………..64
Keeping the Oil and Gas Pipelines Operating as Assets: Safety Issues Can Make
Them Liabilities
Carol Sullivan……………………………………………79
An Analysis of Alternative Work Arrangements to Address the Gender Gap in
Public Accounting
Anderson and Smith………………………………………….87
Variations in Unfunded Pension Liabilities Across U.S. States Paul A. Tomolonis………………………….105
Consistency of Ethical Analysis Among Accounting Students
Flynn, Deno and Buchan………………………………………124.
Fraud-Detecting Effectiveness of Management and Employee Red Flags Aa
Perceived by Three Different Groups of Professionals
Moyes, Anandarajan and Arnold………………………………….133
Journal of Business and Accounting
Volume 12, No 1; Fall 2019
4
OIL RESERVES AND FINANCIAL
INFORMATION DURING 2011 - 2015
Anwar Salimi
University of Laverne
ABSTRACT The price of oil fluctuated wildly from 2011–2015 and fell to about half
the price it was at its peak. This paper investigates whether proven oil reserves of
companies provide information for the market value of the firm. In doing so, we
also examine if proven oil reserves help in predicting future earnings of companies
incremental to other measures such as current earning and book value. The study
uses financial models from the accounting literature that are influential and have
been used in other studies for different purposes and in different industries. The
sample consisted of Oil Exploration and Production firms (GICS subindustry
10102020) plus Integrated Oil firms (GICS subindustry 10102010) as included in
the COMPUSTAT database. The sample included 120 firms. Data points were
generated for a 5 year-period from 2011 to 2015. The results of the data analysis
indicate that proven reserve information does not seem to provide any additional
explanatory power for the market value of equity of firms incremental to the
information provided in financial variable such as earnings, book value, accruals
and cash flows. Also, proven reserve information does not seem to aid in predicting
future earnings over and above the information found in financial variables such
as current year earnings, book value and cash flow and accruals.
Key Words: Oil reserves, information content, market value of firm, future
earnings.
INTRODUCTION Accounting for Oil and Gas companies has had a tumultuous history.
There are two methods that are generally accepted for accounting for exploration
costs for oil. These are the successful efforts method and the full-cost method. The
successful efforts method requires that costs of exploration for oil that do not result
in discovering oil and gas should be expensed in the same period as the incurrence
of the expense. The other method is the full-cost method which requires that
exploration costs for all wells, whether they are successful or not should be
capitalized as assets and expensed when oil is extracted from successful wells.
Both of these methods are used by different companies in the oil industry. The
depletion base is calculated under both methods based on cost of exploration which
includes acquisition cost of the deposit, exploration costs, development costs and
restoration costs. However, for successful efforts firms the depletion base only
includes these costs for successful wells. For full-cost firms the depletion base
includes the costs of both successful wells as well as dry holes.
Journal of Business and Accounting
5
A recounting of the changes in accounting for oil shows the political nature
of standard setting. In 1977 the Financial Accounting Standards Board (FASB)
published a standard that required all oil firms to use successful efforts accounting.
This standard was met with strong opposition by many oil companies in particular
by small companies. These companies felt that being forced to follow the
successful efforts method would cause their earnings to fluctuate widely from year
to year and thus cause investors and creditors to consider these companies more
risky and make it harder and more expensive for them to raise capital. Thus they
argued it would cause them to cut back on new exploration. The oil companies
lobbied extensively in Congress. The Department of Energy at that time also
commented that companies using full-cost at that time would cut back on their
exploration activities if they had to switch to successful efforts. Congress was
swayed by these arguments that the successful-efforts method would result in a
reduction in exploration for oil and increase the dependence of the USA on foreign
countries for oil. The Justice Department asked the Securities and Exchange
Commission (SEC) to postpone uniform adoption of successful efforts accounting
until the SEC could determine if the information reported to investors would be
improved and what the effect of the standard would be on exploration activities.
In 1978 in response to all this negative feedback about the FASB standard,
the SEC examined the issue and concluded that both successful efforts and full
cost accounting had problems. The SEC advocated a different method called
reserve recognition accounting (RRA) in which as soon as a company discovers
oil, it would report the value of the oil in the primary financial statements and
prepare an income statement based on RRA. From 1979 to 1981, because of the
SEC criticism, the FASB rescinded their standard that only permitted successful
efforts accounting and companies were allowed to use full-cost accounting again.
The SEC however encountered major problems in implementing RRA. The
estimation of the reserves and the selling price and discount rate proved to be a
difficult and unreliable task.
In 1981 the SEC abandoned RRA and the requirement that it be used in
the primary financial statements of oil companies because the figures were
considered too unreliable.
The SEC however required reserve information to be disclosed as
supplementary information in the financial reports of oil companies (Accounting
Standards Codification (ASC) 932-235-55-1). One of the most significant items
disclosed in the supplementary disclosures is the amount of proven reserves that
an oil company possesses. This is still however subjective information because the
amount of reserves is hard to estimate and companies are always changing their
estimates of the reserves. The determination of “reasonable certainty” (from the
SEC guidance) is subject to different interpretation by different companies.
In a volatile industry like the oil industry it is hard to determine if the
amount of reserves will have an effect on the market valuation of a company and
also in predicting its future earnings. This uncertainty has been compounded in
recent years because of dramatic fluctuations in the market price of oil. The current
Salimi
6
price of a barrel of oil is approximately half of what it was a few years ago and is
subject to wild and unexpected changes.
It is the purpose of this paper to empirically test from 2011-2015 when oil
prices were particularly volatile whether proven oil reserves of companies provide
some information for the market value of the firm incremental to measures such as
earnings and book value of the firm. Also to see if proven oil reserves help in
predicting future earnings of companies incremental to other measures such as
current earning and book value. The price of oil fluctuated wildly from 2011-2015
and fell to about half the price it was at its peak. Thus proven reserves of oil may
or may not provide significant information for the market. The study uses financial
models from the accounting literature that are influential and have been used in
other studies for different purposes and in different industries.
This study will help in increasing our knowledge about the influence of
non-financial variables and also provide input to accounting policy makers.
LITERATURE SEARCH This study falls generally in the area of non-financial variables and how
they influence financial variables of interest such as market value of equity and
future earnings.
The Financial Accounting Standards Board has expressed an interest in
exploring the value of non-financial information reporting to investors and
creditors (Maines et al. 2002). There is some evidence that some firms voluntarily
disclose non-financial performance measures and also that financial analysts use
non-financial measures in their evaluation of firms (Dempsey et al., 1997).
There has also been interest among accounting researchers about the
relevance of non-financial information for investors and creditors. One of the
qualities that accounting information is expected to possess is relevance to
investors and creditors in making decisions.
The research in the value relevance area indicates that the results depend
on the industry studied. Also of course the non-financial variables that are relevant
are different for different industries. Thus Hughes (2000) finds measures of air
pollution relevant for the electric utility industry while Graham et al. (2002) finds
that page views are relevant for e-tailers.
Dempsey et al. (1997) surveyed 420 financial analysts and reported that
analysts make use of traditional financial statements but also use non-financial
information. The study indicated that financial analysts are interested in these
measures as predictors of financial performance.
Previous research in accounting has provided some documentation of the
value of non-financial measures. Amir and Lev (1996) found that in the cellular
telephone industry, non-financial information such as market penetration was
highly value relevant. They found that because of the peculiarities of this
particular industry, financial information such as earnings by itself were not value
relevant but financial information was relevant when combined with non-financial
information. Rautavuori (2005) examined the value-relevance of wireless
companies’ spectrum licenses and industry specific non-financial information.
The study showed that both financial and non-financial variables were value
Journal of Business and Accounting
7
relevant but the non-financial variables were less significant than financial
variables. Thus, there were mixed results in the wireless telecommunication
industry about the value relevance of financial versus non-financial variables in
impacting a company’s stock price. It was also difficult to determine which non-
financial information may be relevant.
Rajgopal et al. (2001) found that web traffic was an important non-
financial indicator of the market values of Business to Consumer (B2C) Internet
firms. Jorion and Talmor (2001) studied the Internet industry and found that web
traffic, a non-financial measure was more value relevant than conventional
financial measures. However they found that the value relevance of the financial
measures increased with time as the industry matured and the relative importance
of web traffic diminished though it was still significant. They suggested that value
relevance of different variables changed with the stage of life of this industry.
Graham et al. (2002) examined the value relevance of non-financial measures of
internet usage across four Internet industry sectors. They found that net income
had no ability to explain stock prices in this industry. With respect to the non-
financial variables, they found that for e-tailers and content/community firms, page
views had the greatest explanatory power but for service companies, unique users
were the most relevant. However, for infrastructure companies, none of the non-
financial variables were significant. In general however non-financial information
in this study was more relevant than financial information. Bartov et al. (2001)
examined the valuation of internet stocks IPO’s and determined that the financial
and non-financial variables that were significant in the valuation of internet firms
were different from those of non-internet firms. Keating et al. (2003) studied the
downturn in the stock prices of internet companies in spring 2000 and concluded
that financial statement information contributed more to explaining the fall in stock
prices than did non-financial information. Thus, for the internet industry also the
evidence of the value relevance of non-financial measures was mixed. The value
relevance of non-financial measures seems to change with the maturity of the
industry.
Hughes (2000) looked at the stock prices of companies in the Electric
Utility industry and examined whether they are affected by their exposure to future
environmental liabilities. She measured this exposure with a non-financial measure
of pollution namely sulfur dioxide emissions. She found that this non-financial
pollution measure was value relevant and its relevance increased when stringent
environmental regulations were passed in 1990. She found that the industry’s
exposure to environmental liabilities decreased firm’s stock prices by 16 percent
on average. Riley et al. (2003) provided evidence of the value of non-financial
measures in the airline industry. They found that non-financial variables showed
incremental value relevance when included along with financial variables in
explaining stock returns. However, they did not find the reverse to be true. They
did not find that financial variables show incremental explanatory power over non-
financial variables when included in the same model. Yang (2003) examined
whether non-financial patent information is value relevant in market valuation for
biotech firms. He found evidence that non-financial patent information captured
Salimi
8
biotech firms values not currently formally valued by traditional financial
indicators. Shortridge (2004) documented the relationship between a non-financial
measure of successful research and development (R&D) efforts and market
valuation of firms in the pharmaceutical industry. She found that R&D did not
explain stock price in her primary model but did so for firms with successful R&D
efforts. Thus, the market appeared to discriminate between successful and
unsuccessful R&D efforts. Thus, we see evidence for value relevance of non-
financial information in the electricity, airline, biotech, and pharmaceutical
industry.
Blazovich et. al. (2013) studied the effect of “being green” on a firms
financial performance. They found that being green did not significantly impact
financial performance but also did not affect the profitability of the firm.
Lee (2012) found that corporate reputation and non-financial variables are
positively associated to corporate financial performance.
The studies reviewed above generally show that non-financial variables
are value relevant for stock prices. However the evidence regarding the value
relevance of non-financial information is somewhat mixed in some industries.
Also the conclusions to be drawn from these studies are highly industry specific
because the non-financial variables studied are industry specific. Also other factors
such as the maturity of the industry have been shown to be significant in
determining the value-relevance of non-financial variables. The studies reviewed
above also indicate that non-financial performance measures may be value relevant
because they predict future financial performance.
There have also been studies in the oil industry regarding measures other
than historical cost accounting to see their effect on market and financial variables.
Previous studies with the oil industry have indicated that historical cost accounting
by itself may be inadequate in measuring the performance of oil companies. This
may be because there is a lot of uncertainty surrounding oil companies
performance such as the risk of drilling a dry well and the time that elapses between
discovery of oil and sale of oil (Quirin et al. 2000).Also estimates of proven level
of oil fluctuate are uncertain and fluctuate dramatically sometimes with improved
technology. The price of oil also fluctuates dramatically.
There have been studies relating to effect of non-financial variables on
market returns and valuation of oil companies. Cormier and Magnan in a 2002
study found cash flows to be associated with market returns.
Bandyopadhyay (1994) shows that whether a firm uses full cost or
successful efforts accounting determines market reaction to the financial
information released by the firm. Teall (1992) found using Canadian oil company
data that reserve disclosure methods provide incremental information content to
the market over historical cost earnings. Doran 1988 goes into incremental
information content of RRA over historical cost earnings and finds it does have
incremental information content. Alciatore (1993) shows that the value of a firm
may be affected by changes in oil price reserves. Spear (1996) shows that stock
market returns may be affected by changes in reserves. Berry and Wright (1997)
Journal of Business and Accounting
9
however find that some measures of reserves are value relevant while others are
not.
There is an implicit assumption in the studies on non-financial variables
reviewed above. The assumption is that non-financial variables are value relevant
because they predict future financial performance which is what is of value to
investors. There are studies which indicate that non-financial performance
measures can predict future financial performance.
However, the literature reviewed above shows that the conclusions to be
drawn from these studies are highly industry specific because the non-financial
variables studied are industry specific. Also, other factors such as the maturity of
the industry and environmental factors have been shown to be significant in
determining the value-relevance of non-financial variables. The value relevance
and effect on future performance can also change in different periods depending
on external environmental factors such as new laws or penalties or the maturing of
the technology used by an industry (Hughes (2000), Jorion and Talmor (2001),
Graham (2002), Keating (2003) and others). Value relevance and the ability to
enhance and predict future earnings may change with environmental factors
affecting the industry. In the case of the oil industry the volatility in oil prices
combined with the drop in oil prices may affect the relevance of oil reserves for
financial performance.
MODELS USED IN THE STUDY Several influential models in the accounting literature were used in the
study. The reason for using several models was to ensure the results were not
biased because of choice of model.
The following valuation model is based on Ohlson (1995, 2001) and Barth
et al. (1999). I have extended it with the addition of the Proven Reserves of Oil
variable. The idea is to see if the Reserves variable provides any incremental
information about the value of the firm above the other financial variables included
in the model.
Model 1:
MVEt = α0 + β1 EARNt + β2BVEt + β3ACCRt + β4PRESt + ε
Where:
MVE = Market value of Equity at time t.
EARN = Abnormal earnings at time t (This is calculated as earnings before
extraordinary items minus beginning book value multiplied by 12%. The 12%
figure is assumed to be the long run rate of return on equities (Dechow et al. 1999
and Barth et al. 1999). Thus the assumption is that only earnings greater than the
12% return would be significant for the valuation of MVE).
BVE = Book value of the firm at time t
ACCR = Accruals of the company at time t. This is the difference between
income before extraordinary items and operating cash flow. ACCR is included in
the model on the basis of Barth et al. (1999) findings that accruals provide
explanatory power for MVE above the financial variables of EARN and BVE
Salimi
10
PRES = Proven reserves of oil for the company at time t. This is our non-
financial variable which is included in order to test for its explanatory power for
MVE.
Another model used in this study based on Barth et al. (1999) is shown
below.
Model 2
MVEt = α0 + β1 EARNt + β2BVEt + β3CFOt + β4PRESt + ε
Where:
CFO = Operating cash flow from the cash flow statement. Barth et al.
(1999) show that cash flow may have more explanatory power for MVE over
EARN and BV.
All the other variables used in this model have been defined previously
above.
The next valuation model is based on Amir and Lev’s [1996] adaptation
of the Ohlson (1995) model.
Model 3
MVEt = α0 + β1 INCt + β2BVEt + β3PRESt + ε
Where:
INC = earnings before extraordinary items
All the other variables used in this model have been defined previously
above.
The next model is meant to predict next years earnings. They are based on
Barth et al. (1999 and 2001).
Model 4
EARNt+1 = α0 + β1 EARNt + β2 BVEt + β3ACCt + β4PRESt + ε1
Where:
EARN (t+1) = next years abnormal earnings. We are trying to observe the
explanatory power of the independent variables for future earnings.
All the other variables have been defined previously.
The rationale for including the proven reserves of oil in the equation are
based on the assumption that reserves, other things being equal, should add to
explaining the market value of the firm and its future earnings. However whether
they add incremental explanatory power in the presence of the other financial
variables of earnings and book value is to be tested. Also in a period of volatility
in oil prices and a drop in market prices of oil, proven reserves may not add any
value to the company that is not already included in the book value of the company.
DATA AND RESULTS The sample consists of Oil Exploration and Production firms (GICS
subindustry 10102020) plus Integrated Oil firms (GICS subindustry 10102010) as
included in the COMPUSTAT database. The sample came to 120 firms. Data was
gathered for a 5 year period from 2011 to 2015. There was missing data for some
of the firms in some of the years. The total number of data points over the 5 year
period came to 459 data points.
Journal of Business and Accounting
11
Descriptive data for the variables used in the sample are listed in Table 1
below:
Table 1
Descriptive Data for Variables
Mean Median Standard Deviation Count
Price per
share 35.89718519 25.98 33.58081032 459
Earnings
per Share 0.102614379 0.68 8.27256132 459
Book value
per share 22.6041024 16.681 23.53117843 459
Proven
reserves per
share 1549.636311 887.9582647 2299.79407 459
Common
Shares
Outstanding
in millions 597.4992832 176.8 1139.966852 459
Market
value of
equity in
millions 26648.135 3546.588 58592.17936 459
Abnormal
earnings in
millions
-
773.4128173 -89.362 4342.950194 459
Book value
in millions 20637.43719 2114.26 42747.42435 459
Accruals in
millions
-
3429.589044 -737 6162.013796 459
Cash from
operations
in millions 5146.006972 647.099 10197.87371 459
Proved
reserves in
thousands
of barrels 1281152.941 100515 2801788.226 459
The variables have a wide range of values as shown by the standard
deviations. This is also indicated by the fact that the median is usually much less
than the mean. This means that the mean is being dragged upward by the larger
firms. All of this indicates that there are a lot of small Exploration and Production
(E&P) firms in the sample. But there is a big standard deviation for the sample
with many large firms as well.
The price per share has a mean of $35.89 and Standard Deviation of 33.58.
Earnings per share has a mean of $.10 and Standard Deviation of 8.27. Book value
per share has a mean of $22.60 and total book value has a mean of $20.6 billion
Salimi
12
with a median of $2.1 billion and standard deviation of $42.7 million. Market
value of equity has a mean of $26.6 billion, a median of $3.5 billion and standard
deviation of $58.5 billion. Cash from operations has a mean of $5.1 million, a
median of 647 million and a standard deviation of $10.1 billion. Abnormal
earnings have a mean of $773 million but the median is negative $89 million. This
means that many firms in the sample are making less than the 12% of book value
threshold. The accruals have a mean of $3.4 billion indicating earnings are greater
than cash flows for the mean sample but the median is a negative $737 million.
The main information to be gleaned from the descriptive statistics is that
there is a mixture of small and large firms in the sample and the variables have a
big standard deviation.
Table 2 shows the regression results for the market value of equity using
valuation model 1.
Table 2
Market Value of Equity Regressed on Financial Variables and Proven Oil
Reserves
Model: MVEt = α0 + β1 EARNt + β2BVEt + β3ACCRt +β4PRESt + ε
Adjusted R Square 0.885779368
Coefficients P-value
Intercept 2135.053452 0.04461291
EARN 5.10225457 2.16088E-
42
*
BVE 0.962620126 3.71596E-
40
*
ACCR -2.51563186 1.27086E-
09
*
PRES -3.41458E-
05
0.96024303
* indicates significance at the 5% level All variables are as defined previously
The results from Table 2 indicate that the financial variables such as
EARN and BVE are significant and the sign of the coefficients is positive as
expected. The market seems to get information from these financial variables. The
sign of ACCR is negative and significant. This is as expected because the market
seems to discount accounting accruals (difference between accounting earnings
and cash flows). This may be because the market uses cash flow information and
discounts the effect of accruals. This negative coefficient for ACCR is also in
accord with Barth et al. (1999). The sign of the non-financial variable for proven
reserves is negative. However the p-value indicates it is not statistically significant.
The coefficient for PRES is miniscule and the p-value indicates that it may be due
Journal of Business and Accounting
13
just to random fluctuation. The negative coefficient could be interpreted to mean
that reserves don’t add to the value of the firm in a time of fluctuating and falling
oil prices. The statistical insignificance of the variable means that reserves do not
provide any incremental information to the market over and above the financial
variables.
Table 3 shows the regression results for the market value of equity using
valuation model 2.
Table 3
Market Value of Equity Regressed on Financial Variables and Proven Oil
Reserves
Model: MVEt = α0 + β1 EARNt + β2BVEt + β3CFOt + β4PRESt + ε
Adjusted R Square 0.88828248
Coefficients P-value
Intercept 2113.536856 0.043474641
EARN 2.830709556 3.80342E-30 *
BVE 0.638711659 2.61891E-10 *
CFO 2.72927847 7.27901E-12 *
PRES -
0.000399329
0.55566451
* indicates significance at the 5% level The results from Table 3 indicate that the financial variables of EARN and
BVE are positive and significant as before. CFO is also positive and significant
indicating that the market gets additional information from cash flows
incrementally to EARN and BE and that cash flows add value to the firm. PRES
is once again negative and statistically insignificant. Thus the negative value may
indicate that reserves are viewed negatively by the market or the result may just be
due to random fluctuations because it is not statistically significant.
Table 4 shows the regression results for the market value of equity using
valuation model 3.
Table 4
Market Value of Equity Regressed on Financial Variables and Proven Oil
Reserves
Model: MVEt = α0 + β1 INCt + β2BVEt + β3PRESt + ε
Adjusted R Square 0.88659203
Salimi
14
Coefficients P-value
Intercept 3528.339721 0.000158277
EARN 4.157299076 4.14534E-58 *
BVE 0.804704463 4.72303E-49 *
PRES -
0.000167475
0.788988792
* indicates significance at the 5% level It is reassuring that even with this formulation of the model; the results are
quite similar to the results from Tables 2 and 3. The financial variables of EARN
and BVE are positive and significant. The PRES variable is once again negative
and statistically insignificant.
Table 5 shows the regression results for next years earnings using
valuation model 4.
Table 5
Earnings for Next Period Regressed on Financial Variables and Proven
Oil
Reserves
Model: EARNt+1 = α0 + β1 EARNt + β2 BVEt + β3ACCRt + β4PRESt + ε1
Adjusted R Square 0.57168288
Coefficients P-value
Intercept -
416.9220197
0.013747907
EARN 0.926602117 7.4314E-42 *
BVE -
0.055465551
5.0841E-06 *
ACCR -
0.035424434
0.61479803
PRES 0.000179759 0.122693742
* indicates significance at the 5% level The results from Table 5 indicate that EARN is positive and significant in
predicting next years earnings. BE however is negative and significant. This may
be because in the calculation of next years EARN we take earnings before
extraordinary items and subtract 12% multiplied by the beginning of year book
value. Thus the larger the book value the bigger the number that we subtract from
next years EARN. ACCR is negative and insignificant. It appears that accruals
may not provide much information in predicting next years earnings. PRES is once
again insignificant in predicting next years EARN. It does not provide any
incremental explanatory or predictive power in addition to the financial variables
already included in the model.
Journal of Business and Accounting
15
CONCLUSION The results of the data analysis above indicate that proven reserve
information does not seem to provide any additional explanatory power for the
market value of equity of firms incremental to the information provided in
financial variable such as earnings, book value, accruals and cash flows.
Also proven reserve information does not seem to aid in predicting future
earnings over and above the information found in financial variables such as
current years earnings, book value and accruals.
The reason for this may because of the subjective nature of estimating
reserves and changes to them due to changes in technology and other variables.
Also the volatility in the price of oil and its recent dramatic drop may make
reserves not relevant for the market value of equity or the future earnings of a
company.
REFERENCES
Alciatore, M. L. “New evidence on SFAS no. 69 and the components of the
change in reserve value.” The Accounting Review, 68(3), 639.
Amir, E., & Lev, B. “Value-relevance of nonfinancial information: The wireless
communications industry.” Journal of Accounting & Economics, 22(1-3), 3-
30.
Bandyopadhyay, S. P. “Market reaction to earnings announcements of successful
efforts and full cost firms in the oil and gas industry.” The Accounting
Review, 69(4), 657.
Banker, R. D., Potter, G., & Srinivasan, D. “An empirical investigation of an
incentive plan that includes nonfinancial performance measures.” The
Accounting Review, 75(1), 65-92.
Barth, M. E., Beaver, W. H., Hand, J. R. M., & Landsman, W. R. “Accruals, cash
flows, and equity values.” Review of Accounting Studies, 4(3-4), 205-229.
Barth, M. E., Cram, D. P., & Nelson, K. K. “Accruals and the prediction of future
cash flows.” The Accounting Review, 76(1), 27-58.
Bartov, E., Mohanram, P., & Seethamraju, C., “Valuation of internet stocks – An
IPO perspective.” Working paper. New York University.
Salimi
16
Berry, K. T., & Wright, C. J. “Value-relevant reserve quantity disclosures: Oil
reserves versus gas reserves.” Petroleum Accounting and Financial
Management Journal, 16(1), 1-14.
Blazovich, J. L., Smith, K. T., & Smith, L. M. “An examination of financial
performance and risk of environmentally friendly 'green' companies.”
Journal of Legal, Ethical and Regulatory Issues, 16(1), 121-134.
Cormier, D., & Magnan, M. “Performance reporting by oil and gas firms:
Contractual and value implications.” Journal of International Accounting
Auditing & Taxation, 11(2), 131-153.
Dechow, P. M., Hutton, A. P., & Sloan, R. G. “An empirical assessment of the
residual income valuation model.” Journal of Accounting & Economics,
26(1-3), 1-34.
Dempsey, S. J., Gatti, J. F., D Jacque Grinnell, & Cats-Baril, W. L. “The use of
strategic performance variables as leading indicators in financial analysts'
forecasts.” Journal of Financial Statement Analysis, 2(4), 61-79.
Doran, B. M., Collins, D. W., & Dhaliwal, D. S. “The information of historical
cost earnings relative to supplemental reserve-based accounting data in the
extractive petroleum industry.” The Accounting Review, 63(3), 389.
Feltham, G. A., & Ohlson, J. A. “Valuation and clean surplus accounting for
operating and financial activities.” Contemporary Accounting Research,
11(2), 689.
Graham, C. M., Cannice, M. V., & Sayre, T. L. “The value-relevance of financial
and non-financial information for internet companies.” Thunderbird
International Business Review, 44(1), 47-70.
Hughes, K. E.,II. “The value relevance of nonfinancial measures of air pollution
in the electric utility industry.” The Accounting Review, 75(2), 209-228.
Ittner, C. D., & Larcker, D. F. “Are nonfinancial measures leading indicators of
financial performance? an analysis of customer satisfaction.” Journal of
Accounting Research, 36, 1-35.
Kallapur, S., & Kwan, S. Y. S. “The value relevance and reliability of brand
assets recognized by U.K. firms.” The Accounting Review, 79(1), 151-172.
Journal of Business and Accounting
17
Keating, E. K., Lys, T. Z., & Magee, R. P. “Internet downturn: Finding valuation
factors in spring 2000.” Journal of Accounting & Economics, 34(1-3), 189-
236.
Lambert, R. A. “Customer satisfaction and future financial performance
discussion of are nonfinancial measures leading indicators of financial
performance? an analysis of customer satisfaction.” Journal of Accounting
Research, 36, 37-46.
Lee, J., & James Jungbae Roh. “Revisiting corporate reputation and firm
performance link.” Benchmarking, 19(4/5), 649-664.
doi:http://dx.doi.org/10.1108/14635771211258061
Maines, L. A., Bartov, E., Fairfield, P. M., Hirst, D. E., & et al.
“Recommendations on disclosure of nonfinancial performance measures.”
Accounting Horizons, 16(4), 353-362.
Nagar, V., & Rajan, M. V. “The revenue implications of financial and
operational measures of product quality.” The Accounting Review, 76(4),
495-513.
Ohlson, J. A. “Earnings, book values, and dividends in equity valuation.”
Contemporary Accounting Research, 11(2), 661.
Ohlson, J. A. “Earnings, book values, and dividends in equity valuation: An
empirical perspective.” Contemporary Accounting Research, 18(1), 107-
120.
Quirin, J. J., Berry, K. T., & O'Bryan, D. “A fundamental analysis approach to
oil and gas firm valuation.” Journal of Business Finance & Accounting,
27(7/8), 785-820.
Rajgopal, S., Kotha, S., & Venkatachalam, M., “The relevance of web traffic for
stock prices of internet firms.” Working paper. University of
California,Irvine
Rautavuori, J. “Spectrum licenses and industry-specific nonfinancial
information.” Working paper. University of Helsinki.
Rebecca Toppe Shortridge. “Market valuation of successful versus non-
successful R&D efforts in the pharmaceutical industry.” Journal of Business
Finance & Accounting, 31(9/10), 1301-1325.
Salimi
18
Riley, R. A.,Jr, Pearson, T. A., & Trompeter, G. “The value relevance of non-
financial performance variables and accounting information: The case of the
airline industry.” Journal of Accounting and Public Policy, 22(3), 231-254.
Teall, H. D. “Information content of Canadian oil and gas companies' historic
cost earnings and reserves disclosures.” Contemporary Accounting
Research, 8(2), 561.
Yang, Y., “The value-relevance of non-financial information: The biotechnology
industry.” Ph.D. Dissertation, University of Tennessee, Knoxville.
Journal of Business and Accounting
Vol 12, No 1; Fall 2019
19
DOES THE UNITED STATES POSTAL SERVICE
CHARGE AMAZON TOO LITTLE?
Thomas L. Barton
John B. MacArthur
University of North Florida
ABSTRACT: A concern has been raised by the President of the United States
about the United States Postal Service (USPS) charging significantly discounted
prices on the packages it ships for Amazon.com, Inc. Also, there were other critics
of USPS’s special package delivery pricing practices for Amazon, United Parcel
Service (UPS), and Federal Express Corporation (FedEx). This paper describes
how the USPS uses very sophisticated cost accounting, statistical, and other
approaches to make sure its competitive products cover all their relevant costs and
make an appropriate contribution towards covering institutional common costs.
Market conditions are carefully considered in determining the prices of
competitive products, including providing volume and other discounts to high
volume customers, such as Amazon, UPS, and FedEx. It does seem that the
concern expressed over the pricing of USPS competitive products is more
appropriately directed towards the declining market for its market-dominant
products.
Key Words: United States Postal Service; relevant costs; pricing decisions;
customers; competitors.
INTRODUCTION
A concern has been raised by the President of the United States about the United
States Postal Service (USPS) charging significantly discounted prices on the
packages it ships for Amazon.com, Inc. (Amazon) (e.g., Lombardo & Ziobro,
2017). It was later reported that the President continued his criticism of Amazon
using the USPS as its “delivery boy” (e.g., Taylor, 2018a, 2018b). There has been
considerable pushback on the validity of the basis for such criticism of the
arrangement between USPS and Amazon (e.g., Atkinson, 2018; Save the Post
Office, 2018a).
The criticism of the package delivery arrangement between the USPS and Amazon
may have been at least partly stimulated by the then 10 and, more recently, “11
consecutive years of net losses” reported by the USPS (United States Postal
Service, 2019). For example, net losses of $2,742 and $3,913 million were reported
by the USPS for fiscal years ended September 30, 2017, and 2018, respectively
(United States Postal Service, 2018c, 16). As Megan Brennan, Postmaster General
and CEO stated: “The Postal Service receives no dollars for operating expenses
Barton and MacArthur
20
and relies on the sale of postage, products and services to fund its operations”
(Brennan, 2017). As shall be pointed out below, there were other critics of USPS’s
special package delivery pricing practices for Amazon as well as for its two other
big customers, United Parcel Service (UPS) and Federal Express Corporation
(FedEx).
Like private organizations, the USPS should consider costs, customers, and
competitors in making its product pricing decisions (e.g., Datar & Rajan, 2018,
525-526), as well as pertinent federal government statutes and regulations that
restrict its freedom in making these decisions. As an independent agency of the
United States Federal Government, pricing by the USPS is regulated by federal
legislation and the Postal Regulatory Commission.
The USPS has two types of products, market-dominant and competitive products.
The USPS stated the following about these two product categories (United States
Postal Service, 2019):
First-Class Mail, Marketing Mail, and Periodicals are the major categories
of Market-Dominant mail. Altogether, Market-Dominant mail accounted
for approximately 70 percent of total revenue in 2017. These products are
subject to a strict price cap tied to changes in the Consumer Price Index,
restricting our ability to generate revenue.
Our Package business accounts for the remainder of our revenue, and we
compete for customers with other package delivery providers in the open
market. Package products must cover their costs and are therefore subject
to a price floor. In addition, by law, the competitive portion of these
products as a whole must contribute at least 5.5 percent toward
institutional costs to help fund our universal service obligation. In fact,
they contribute far more than 5.5 percent and this level of contribution is
currently under review by the Postal Regulatory Commission.
It is the package/parcel business services provided by the USPS for Amazon in the
competitive category that has been heavily criticized and is the focus of this paper.
Unusually, two of the USPS’s major competitors for parcel delivery, UPS and
FedEx, are also major customers of the USPS by relying on it “for the back-end of
their two-to-seven-day delivery options” (Stevens, 2014). The USPS has a
competitive advantage over UPS and FedEx as it “is the only organization in the
country that has the resources, network infrastructure and logistical capability to
regularly deliver to every residential and business address in the nation” (United
States Postal Service, 2018b, #3).
In another strange twist, the USPS is also a customer of UPS and FedEx where
they have a competitive advantage. The USPS describes its complex relationship
with FedEx and UPS in the following way (United States Postal Service, 2018b,
#5):
Journal of Business and Accounting
21
The Postal Service both competes and collaborates with the private sector.
UPS and FedEx pay the Postal Service to deliver hundreds of millions of
their ground packages, and USPS pays UPS and FedEx for air
transportation.
Next, the three elements of pricing decisions, costs, customers, and competitors,
will be examined in turn.
COSTS
The USPS costs listed in its published operating results appear to be mainly fixed.
For example, Exhibit 1 reproduces the operating results reported by the USPS for
fiscal years ended September 30, 2017, 2016, and 2015 (United States Postal
Service, 2017, 6). The operating expenses are mostly associated with USPS’s
employees. Employee related compensation and benefits are subject to the USPS’s
“collective bargaining obligations and our obligation to participate in federal
benefits programs” (United States Postal Service, 2017, 22) that would severely
restrict the ability of the USPS to reduce such costs in the short run. It appears from
United States Postal Service (2017, 21), that “All other operating expenses” shown
in Exhibit 1 include depreciation, supplies and services, and rent and utilities,
which are likely to be largely fixed costs.
Exhibit 1: United States Postal Service Operating Results for Fiscal
Years Ended September 30, 2017, 2016, and 2015 (Dollars in Millions)1
FY2017 FY2016 FY2015
Operating results
Total revenue $69,636 $71,498 $68,928
Operating expenses
Compensation and benefits2 49,108 48,441 47,278
Unfunded retirement benefits 2,658 248 241
Retiree health benefits 4,260 9,105 8,811
Workers’ compensation (797) 2,682 1,760
Transportation 7238 6,992 6,579
All other operating expenses 9,743 9,431 9,157
Total operating expenses3 72,210 76,899 73,826
Loss from operations (2,574) (5,401) (4,898)
Investment & interest income-net (168) (190) (162)
Net Loss (2,742) (5,591) (5,060) 1. Source: United States Postal Service (2017, 6).
2. “Excludes amortization of unfunded retirement benefits, retiree health
benefits and workers’ compensation” (United States Postal Service, 2017,
6, footnote 1).
3. Total operating expenses were added by the authors and were not in the
original.
Barton and MacArthur
22
Taking a long enough time horizon, fixed costs can be reduced. For example, the
USPS reported in January 2019 that it had undertaken aggressive network and
infrastructure rightsizing actions that had led to “cost cutting, efficiency
improvements, and targeted innovation that resulted in approximately $13.4 billion
in annual savings through 2017 […] by consolidating 363 mail processing facilities
and 29,000 delivery routes; modifying retail hours at more than 13,000 Post
Offices; reducing our total workforce size by more than 152,000 through attrition;
negotiating contracts that control wages and benefits and increased workforce
flexibility; and through reductions in administrative overhead” (United States
Postal Service, 2019). The achievement of workforce reduction through its
rightsizing efforts illustrates the ability of the USPS to significantly reduce its
number of employees and its employee related compensation and benefits over the
longer-term through attrition and other ways. Because all costs are variable given
a sufficient time horizon, the term long-term variable costs is sometimes used
instead of the more commonly used term of fixed costs.
In its “FY2018 Performance Plan” section, the United States Postal Service (2017,
23) stated the following about “compensation and benefits” under the subheading
of Controllable expenses:
Compensation and benefits expenses are planned to increase by $0.2
billion, primarily due to contractually required wage increases. The impact
of these wage increases will be mitigated by a larger portion of new, less
expensive employees. We also plan to reduce the number of work hours
through increased efficiency.
Presumably, the “new, less expensive employees” that were planned to be hired by
the USPS would include replacements for USPS employees who were expected to
retire or leave the USPS for other reasons during FY2018.
Given the USPS has mainly fixed costs, it might be expected that its incremental
costs of delivering Amazon’s and other companies’ pre-sorted packages along with
mail is likely to be relatively small. This USPS ‘last-mile’ service used by Amazon
and other companies is called ‘Parcel Select’ (Lombardo and Ziobro, 2017). In
response to critics who several years before the President had questioned the
sufficiency and transparency of the USPS parcel delivery prices for its three big
customers, UPS, FedEx, and Amazon, a former Postmaster General, Patrick R.
Donahoe, was reported to have stated: “[…] the criticism is unwarranted. ‘We
make money on it. We wouldn’t do it if we didn’t make money on it […]. Postal
carriers are already delivering to every mailbox, so transporting the presorted
packages doesn’t add significantly to costs […]’” (Stevens, 2014). Also, in
response to a later critical article, Joseph Corbett, the USPS’s CFO stated: “By law
our competitive package products, including those we deliver for Amazon, must
cover their costs” (Corbett, 2017). However, as the author of the critical article
stated: “But with a networked business using shared buildings and employees,
calculating cost can be devilishly subjective” (Sandbulte, 2017). Therefore, it is no
Journal of Business and Accounting
23
surprise that the USPS uses complex procedures, including activity-based costing
(ABC), to ensure that the prices of its competitive and market-dominant products
are in compliance with the extant regulations (Office of Inspector General United
States Postal Service, 2013).
USPS uses Activity-Based Costing The Code of Federal Regulations (CFR), Title 39, section 3633(a) (1) forbids the
cross subsidization of competitive postal products by market-dominant postal
products that means the revenues of competitive products must cover their costs
(Postal Regulatory Commission, 2016a, 17). The USPS adopted an ABC system
that was developed by private-sector companies such as Deere & Company (that
“coined the name ‘activity-based costing’”) to more accurately assign overhead
costs to products for competitive bidding and other purposes (MacArthur, 1992,
37). As discussed in cost/management accounting textbooks, a major benefit from
implementing ABC systems is the correction of product-cost cross-subsidization
that results from some products being overcosted and other products being
undercosted--often caused by overcosted high-volume products subsidizing
undercosted low-volume products--using traditional costing systems (e.g., see
Datar and Rajan, 2018, 152-196). The USPS conducted an ABC pilot-study during
2002-2003 (United States Postal Service, 2002) and Coopers and Lybrand used
ABC in a credit and debit card payment program feasibility study that USPS hired
them to conduct during the mid-1990s (Carter, et. al., 1998). Incidentally, UPS also
uses an ABC system (Activity Based Costing at UPS, No Date).
The USPS uses four steps to determine the costs attributable to its products that
are described in the following way by the Postal Regulatory Commission (2016a):
1. Divide Costs Among Segments and Components
Costs are divided up into components for analysis. Accounting costs are
first divided among 18 cost segments. The segments are then further
divided into identifiable cost components and then into elements
representing a discrete activity; there are over a hundred components and
many more elements.
2. Identify a Cost Driver and Find Volume-Variable Costs
For each cost element, a cost driver is identified that reflects the essential
activity of that element. For example, the cost driver for Inter-NDC
highway transportation is cubic-foot-miles. The volume-variable cost pool
is then found by using the relationship between the element’s cost and its
cost driver. The relationship is first used to create volume-variable costs
according to methods described below.
3. Distribute Costs to Products
After the pool of volume-variable cost for each product is determined, it
is distributed to the various products. These methods of distribution are
also discussed below.
Barton and MacArthur
24
4. Calculate Unit Volume-Variable Cost
Total volume-variable cost for each product is determined by summing the
volume-variable costs for that product across components. Unit volume-
variable costs are then found by dividing a product’s total volume-variable
costs by its originating volume.
The Code of Federal Regulations, Title 39, section 3622(c)(2) mandates that the
attribution of costs to Postal Service products must be through “reliably identified
causal relationships” (Postal Regulatory Commission, 2016a, 14). Order No. 3506
makes it clear that the determination of short-run variability of costs “is determined
in a variety of ways: econometric modelling, expert judgment, costing systems,
etc. […] If a component has no variability, it has no relationship with volume and
is considered institutional” (Postal Regulatory Commission, 2016a, Appendix A,
15). Because of economies of scale and economies of scope (the variety of postal
products delivered together), the marginal cost per unit of postal products declines
with increasing unit quantities and varieties of postal products that are delivered
together (Postal Regulatory Commission, 2016a, 35).
For example, Exhibit 2 reproduces Figure A-10 in Order No. 3506 with three
illustrative postal products (Postal Regulatory Commission, 2016a, Appendix A,
20). The variable costs per unit for all three hypothetical products is $1 (the
marginal cost of the last unit) and the total variable costs are represented by the
rectangle area immediately above the horizontal axis from 1 to 20 units that is
shaded blue when color is shown. The area immediately above the rectangle area
bounded on the top by a downward sloping marginal cost curve that is shaded
green when color is shown, represents the inframarginal costs that decline from $5
to $1 with increasing units due to economies of scale and economies of scope of
these products. As the number of postal products increase, the marginal cost per
unit should become a stable number at higher volumes ($1 in the hypothetical
example) (e.g., see Save the Post Office, 2018b). Before Order No. 3506,
inframarginal costs were not included in incremental product costs but were
considered to be 100 percent institutional costs. In Order No. 3506, the Postal
Regulatory Commission (2016a) discussed three proposals made by the UPS that
included a reasoned argument for all inframarginal costs to be classified as variable
costs so that they would be included in the costs USPS competitive products prices
should collectively have to cover. The Postal Regulatory Commission rejected the
reasoning provided by the UPS but in its analysis identified “[…] additional costs
that are reliably identified and causally related but not currently attributed, and
represent a significant improvement over the current methodology. In this Order,
the Commission adopts the incremental cost methodology to determine
attributable costs. This results in the attribution of those inframarginal costs that
meet the statutory requirements for cost attribution” (Postal Regulatory
Commission, 2016a, 3). These attributable inframarginal costs are illustrated next
with the aid of Exhibit 2.
Journal of Business and Accounting
25
Exhibit 2: Marginal Cost Curve with Products Grouped Together1
1 Exhibit 2 is a copy of “Figure A-10: Marginal Cost Curve with Products
Grouped Together” from Postal Regulatory Commission (2016a, Appendix
A, 20). Earlier Figures A-1 through A-9 show “Cost $” and dollar currency
is clearly meant for Figure A-10, too.
_________________________________________________________________
The inframarginal costs attributable to the hypothetical third Postal Service
product in Exhibit 2 is the small area under the sloping marginal cost curve from
unit 14 through unit 20 that is above the black horizontal line that represents the
marginal cost of the last unit that is $1. Each of the three hypothetical Postal
Service products would take their turn as the third product to determine their
attributable inframarginal costs in the same manner (Postal Regulatory
Commission, 2016a, Appendix A, 19).
On the same date as Order No. 3506, Order No. 3507 was also published by the
United States Postal Regulatory Commission to give “notice of proposed
rulemaking on changes concerning attributable costing” that were discussed in
Order No. 3506 and illustrated above (Postal Regulatory Commission, 2016b).
Order No. 3507 invited written comments from interested parties within a 30-day
window on the proposed rule changes to attributable costing. Subsequently, Order
No. 3641 reported that comments were received from Amazon Fulfilment
Services, Inc., the Public Representative, the Postal Service, UPS, and the Valpak
Franchise Association, Inc. (Postal Regulatory Commission, 2016c, 3). Some
modifications were made to the final rules in response to the written comments.
Among other changes, the Postal Regulatory Commission revised CFR, Title 39,
Section 3015.7b to be as follows (Postal Regulatory Commission, 2016c, 11):
6
5
4
3
2
1
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Unit of Cost Driver
Mar
gina
l Cos
t
Barton and MacArthur
26
(b) Each competitive product must recover its attributable costs as defined
in 39 U.S.C. 3631(b). Pursuant to 39 U.S.C. 3631(b), the Commission will
calculate a competitive product's attributable costs as the sum of its
volume-variable costs, product-specific costs, and those inframarginal
costs calculated as part of a competitive product's incremental costs.
Product-specific costs include advertising for one particular USPS product only
(Postal Regulatory Commission, 2016a, footnote 12, 9).
Order No. 3641 stated that the final rules were adopted by the Postal Regulatory
Commission despite a request by UPS in its Comment to postpone implementation
because Order No. 3506 was currently being reviewed by the Court (Postal
Regulatory Commission, 2016c). The decision by the Commission to not defer
implementation of the rule changes was validated as it was later reported that the
UPS was unsuccessful in appealing Order No. 3506 in the D.C. Circuit Court of
Appeals (Save the Post Office, 2018b, 2018c, 2018d).
Pricing by USPS In addition to covering the attributable costs discussed above, “Section 3633(a)(3)
of Title 39 of the United States Code requires the Commission to ‘ensure that all
competitive products collectively cover what the Commission determines to be an
appropriate share of the institutional costs of the Postal Service’” that historically
was determined to be a 5.5 percent markup (Federal Register, 2019). Institutional
costs are defined by the USPS as: “Postal costs that cannot be directly or indirectly
assigned to any mail class or product. They can be considered common costs or
overhead costs needed for overall operations” (United States Postal Service,
2018a). Revised rules for determining the appropriate percentage of institutional
costs were published in January 2019 in Order No. 4963 and an increased
percentage of 8.8 percent was determined using a new “formula-based approach”
for Fiscal Year 2019 (Postal Regulatory Commission, 2019a, 2019b).
Interestingly, in Order No. 3506, the Postal Regulatory Commission (2016a, 61)
stated:
Using incremental costs for attribution need not alter the Postal Service’s
pricing strategy, nor should it. While each product’s attributable cost will
be equal to its incremental cost, marginal costs should remain the Postal
Service’s basis for setting prices, with the application of appropriate
markups to ensure each product covers its incremental costs and provides
an appropriate share of institutional costs. Effectively, the average price of
a product should meet or exceed the product’s average incremental cost
(the incremental cost divided by the number of pieces). This would result
in products having a cost coverage of 100 percent or greater.
This statement is consistent with the recognition that marginal costs, plus any
markups required by regulations to, inter alia, prevent cross-subsidization of
competitive products by market-dominant products, to be consistent with costs
setting the floor and not the ceiling on competitive product prices. Higher prices
Journal of Business and Accounting
27
can be set for competitive products if these can be justified by market conditions
that include expected customer and competitor reactions to higher prices.
However, it has been pointed out that profit-maximization is not the goal of USPS
(Atkinson, 2018, 10). If raising package prices to maximize revenues and profits
from competitive products also leads to lower demand, this likely hurts US citizens
(customers), particularly those with lower incomes and those living in remote
areas, violating USPS’s regulations (Atkinson, 2018, 10).
The discussion in Order No. 4963 (Postal Regulatory Commission, 2019a) makes
it very clear that the most recent decision to increase the markup of competitive
market parcel delivery products from 5.5 percent to 8.8 percent to increase the
contribution of competitive products towards institutional costs, was made
following very careful consideration of market conditions. For example, Table II-
2 (page 7) presented the USPS and its combined competitor’s overall parcel
delivery market growth by revenue from FY 2007 through FY2017. Additionally,
Table II-3 (page 8) showed the increasing parcel delivery market share by revenue
gained by the USPS from FY 2007 (9.22 percent) through FY2017 (19.36 percent)
versus a decline in parcel delivery market share by revenue for its combined
competitors from FY 2007 (90.78 percent) through FY2017 (80.64 percent). Also,
Table II-4 (page 11) showed that the average price increased every calendar year
for the USPS, UPS, and FedEx from 2008 through 2019, along with the steadily
increasing market size.
The growth in revenue from USPS’s competitive products contrasts with the
continually declining volume and revenue from its market-dominant mail products
with increasing competition from growing e-commerce (Atkinson, 2018, 1). For
example, the United States Postal Service (2018c, 18) reported the following
(italics in the original):
“[The] combined revenue from First-Class Mail and Marketing Mail
declined by $827 million for the year ended September 30, 2018,
compared to the prior year. This decline was attributable to a decline in
combined volume of nearly 3.2 billion pieces. These declines in First-
Class Mail and Marketing Mail revenue were more than offset by the
increase in Shipping and Packages revenue of nearly $2.0 billion, or
10.1%, as we continued to experience growth in this lower-contribution
service category throughout 2018.
The competitive products’ combined surplus above cost are reducing the losses
caused by the declining market-dominant mail that should be the focus of concern
(Atkinson, 2018, 11). More specifically, USPS’s customers and competitors for
its competitive products are separately considered next in turn.
CUSTOMERS
In addition to published prices for parcel delivery to retail customers, the USPS
negotiates special lower-price agreements that are unpublished, called “negotiated
Barton and MacArthur
28
service agreements” (NSAs), for major customers who offer high volume
opportunities for parcel delivery using the USPS’s competitive advantage “in last-
mile business-to-consumer delivery of lightweight parcels” (Postal Regulatory
Commission, 2019a, 8-9). Along with the UPS and FedEx, Amazon has an
unpublished NSA with the USPS for last-mile deliveries, “with Amazon’s own
transportation and distribution network performing the upstream work” (Postal
Regulatory Commission, 2019a, 10). Rather than hurting the USPS financially, its
NSA agreements with its competitors called “‘co-opetition’ results in lower costs
for each of the three competitors, which in turn lowers the overall cost of delivering
parcels and improves the productive efficiency of the market” (Postal Regulatory
Commission, 2019a, 10). For example, USPS does not have to incur the costs of
sorting packages as Amazon “has a network of more than 20 ‘sort centers’ where
customer packages are sorted by zip code, stacked on pallets and delivered to post
offices for the final leg of delivery” (Sink and Soper, 2017). Interestingly, it was
stated in United States Postal Service (2018c, 1) that: “No single customer
represented more than 6% of operating revenue for the years ended September 30,
2018, 2017, and 2016.” It might be speculated that Amazon is the customer that
represents about 6% of USPS’s operating revenue.
It was reported that Amazon’s 2017 prime member shipments were more than 5
billion items and that Prime members in the United States have increased to 90
million (Perez, 2017). Also, it was reported by Sink and Soper (2017) that “David
Vernon, an analyst at Bernstein Research Bernstein […] estimated in 2015 that the
USPS handled 40 percent of Amazon’s volume in 2015” that is a significant
volume of parcels and certainly merits a volume discount. That is a ‘lot of eggs in
one basket’ and losing this dominance to competitors and/or to Amazon insourcing
more of its transportation needs would likely lose USPS significant positive
contribution margin that probably would not be matched by commensurate savings
in institutional overhead costs.
On August 10, 2018, the USPS announced proposed increased prices from January
27, 2019, for both its market-dominant services, such as First-class Mail, and its
competitive services, including Parcel Select that is used by Amazon and other
companies for the ‘last mile’ delivery of products (Ziobro, 2018). These price
increases were approved by the Postal Regulatory Commission on November 13,
2018 (Postal Regulatory Commission, 2018). A significant postal increase might
indeed lead to Amazon insourcing more deliveries. For example, it was earlier
reported that:
A sudden increase in postal services rates would cost Amazon about $2.6
billion a year, according to an April report by Citigroup. […] Amazon has
been setting up its own shipping operations in the U.S. and elsewhere in
the world to minimize costs. […] Amazon is experimenting with a new
delivery service of its own that is expected to see a broader roll-out in this
coming year. Under the program, Amazon would oversee the pickup of
Journal of Business and Accounting
29
packages from warehouses of third-party merchants and delivery to home
addresses (Sink and Soper, 2017).
A later published article commenting on the proposed 2019 increases in Postal
Service shipping and mailing prices, reported an estimate by Credit Suisse that
they would cost Amazon greater than $1 billion in 2019 (Frank, 2018).
It may therefore not be surprising that Amazon is increasingly using its own trucks
and drivers for package deliveries to its customers because its “shipping costs have
exploded, nearly doubling from 2015 to 2017, to $21.7 billion” (Peterson, 2018)
that likely will be exacerbated by USPS’s most recent price increases that evidently
raised the NSA agreement rates charged for the parcel services provided by the
USPS for Amazon and other companies (Ziobro, 2018). Herrera and Qian (2019)
reported that “Amazon now delivers nearly half of its orders, compared with less
than 15% in 2017, according to estimates from research firm Rakuten
Intelligence.” Also, Amazon announced that it is planning to purchase 100,000
electric vehicles over a ten-year period from 2021 through 2030 for package
delivery to customers and to support its goal “to be carbon-neutral by 2040”
(Thomas, 2019). This movement towards delivering its own parcels to customers
does not support the notion that Amazon management believes it is getting a “real
deal” on its delivery costs provided by third parties such as USPS, UPS, FedEx,
and other delivery partners.
COMPETITORS
FedEx Corporation, DHL International GmbH, and UPS are identified by D&B
Hoovers (2018) as the three leading competitors of the USPS. However, the Postal
Regulatory Commission (2019a, 8) states that “the parcel delivery market has three
primary competitors that make up the majority of the market: United Parcel
Service, Inc. (UPS), Federal Express Corporation (FedEx), and the Postal
Service.” In some ways the three competitors complement each other with
different areas of specialization. The Postal Regulatory Commission (2019a, 8-9)
describes this “firm specialization and product differentiation” competition in the
following way:
One of the primary ways these three competitors have competed with one
another for customers is through firm specialization and product
differentiation. Each of these firms has developed specialties in certain
types of delivery: FedEx specializes in international and express delivery;
UPS specializes in business-to-business delivery; and the Postal Service
specializes in last-mile business-to-consumer delivery of lightweight
parcels. As a result, each of these competitors offers products with
differences in a range of features, including price, service, reliability,
tracking features, and the availability of ancillary services such as
insurance and return options.
However, the UPS has been aggressive in trying to get the USPS to raise its
competitive product prices, presumably to enable the UPS to be more competitive
Barton and MacArthur
30
in it pricing but this would likely also result in the UPS having to pay the USPS
increased rates for postal services provided to the UPS in any renegotiated NSA.
Order No. 3506 was the response by the Postal Regulatory Commission (2016a)
to a petition filed by the UPS with its “proposed changes to postal service costing
methodologies (UPS proposals one, two, and three)” that would lead to an increase
in the costs that competitive products collectively must cover. Because the Postal
Regulatory Commission denied the UPS proposals in Order No. 3506, the UPS
“petitioned the D.C. Circuit Court of Appeals for review” that was rejected “by a
three-member panel of the Court” and the “UPS filed a new petition, this one for
a ‘rehearing en banc,’ asking the full Court to reconsider the panel’s earlier
decision” (Save the Post Office, 2018c). As reported above, this appeal was also
unsuccessful (Save the Post Office, 2018d).
The United States Postal Service (2018c, 5) reported increasing competition for
‘last-mile’ parcel delivery from major existing customers in the following way
(hyphen inserted in ‘Wal-Mart’):
The primary competitors of our Shipping and Packages services are FedEx
Corporation and United Parcel Service, Inc., as well as other national,
regional and local delivery companies and crowdsourced carriers. We see
additional competition coming from existing customers who are testing
and in some cases implementing “last-mile” services. These companies
include Amazon.com, Inc.; Wal-Mart, Inc. and Target Corporation. […]
The growth in our Competitive services revenues over the past five years
is largely attributable to three major customers. For the years ended
September 30, 2018, 2017 and 2016, combined revenue from our three
largest customers (excluding mail service providers) represented
approximately 8.3%, 7.6% and 5.8% of operating revenue, respectively.
The growth in our Competitive service revenues over the past five years is
largely attributable to these three customers. Each of these customers is
building delivery capability that would enable them to divert volume away
from us over time.
Increasing major competition in its traditional area of competitive advantage, ‘last-
mile’ delivery, must be of great concern to USPS management and the Postal
Regulatory Commission. In fact, it was reported that the USPS delivered less
packages (-3.2 percent), albeit with an increase in revenue (4.8 percent), in the
quarter ended June 30, 2019, because Amazon, UPS, FedEx, and other companies
delivered an increased number of online ordered packages to homes (Ziobro,
2019). However, FedEx decided to end both its domestic air shipping contract and
ground package delivery contract with Amazon during summer 2019, while
continuing to ship internationally for Amazon (Ziobro and Mattioli, 2019). This
should increase opportunities for the USPS and other shipping organizations to
deliver more packages for Amazon.
Journal of Business and Accounting
31
CONCLUSIONS
The USPS uses very sophisticated cost accounting, statistical, and other
approaches to make sure its competitive products cover all their relevant costs and
make an appropriate contribution towards institutional common costs. Market
conditions are carefully considered in determining the prices of competitive
products, including providing volume and other discounts to high volume
customers, such as Amazon, UPS, and FedEx, for ‘last mile’ deliveries. The United
States Postal Service (2018c, 25) states that the Parcel Select ‘last-mile’ deliveries
of pre-sorted packages are among the lowest-priced services because of
competitive pricing but still generate a positive contribution margin, albeit
relatively lower than for other package delivery services. It is vitally important that
the USPS continues to be competitive in its pricing for ‘last mile’ parcel services
or it will quickly lose significant sales volume that generates a positive
contribution margin. It does seem that the concern expressed over the pricing of
USPS competitive products is more appropriately directed towards the declining
market for its market-dominant products.
REFERENCES
Activity Based Costing at UPS (No date). Accessed February 18, 2019,
http://govinfo.library.unt.edu/usps/offices/domestic-
finance/usps/pdf/Holsen.pdf.
Atkinson, R. D. (2018). Why the Postal Service is Not Subsidizing Package
Delivery. Information Technology & Innovation Foundation, June 18,
“Read Report” link at: https://itif.org/publications/2018/06/18/why-
postal-service-not-subsidizing-package-delivery/.
Brennan, M. J. (2017). Letter from the Postmaster General. United States Postal
Service FY 2017 Annual Report to Congress,
https://about.usps.com/who-we-are/financials/annual-reports/fy2017.pdf.
Carter, T. L., et al. (1998). How ABC changes the post office.” Management
Accounting [USA], February 1998, p.28+. Gale Academic OneFile,
http://link.galegroup.com/apps/doc/A20509751/AONE?u=jack91990&si
d=AONE&xid=56f3be99. Accessed February 20, 2019.
Corbett, J. (2017). USPS CFO responds to ‘Why the Post Office Gives Amazon
Special Delivery’ article. PostReporter.com, August 2,
http://www.postal-reporter.com/blog/usps-cfo-responds-to-why-the-post-
office-gives-amazon-special-delivery-article/.
D&B Hoovers (2018). United States Postal Service: Competitor Profiles. Dunn
& Bradstreet, Inc. Accessed January 24, 2018,
http://www.hoovers.com/company-
information/cs/competition.United_States_Postal_Service.f741959418bd
451f.html.
Datar, S. M. & Rajan, M. V. (2018). Horngren’s Cost Accounting. 16th Edition.
Hoboken, NJ: Pearson.
Barton and MacArthur
32
Federal Register (2019). Competitive Postal Products: A Rule by the Postal
Regulatory Commission on 01/31/2019, Federal Register, January 31,
https://www.federalregister.gov/documents/2019/01/31/2019-
00399/competitive-postal-products.
Frank, T. (2018). Higher postal fees will cost Amazon more than $1 billion in
2019, estimates Credit Suisse. CNBC.com, October 22,
https://www.cnbc.com/2018/10/22/usps-fees-to-cost-amazon-more-than-
1-billion-in-2019-credit-suisse.html.
Herrera, S. and Qian, V. (2019). How Amazon’s Shipping Empire Is Challenging
UPS and FedEx. The Wall Street Journal, August 29,
https://www.wsj.com/articles/how-amazons-shipping-empire-is-
challenging-ups-and-fedex-
11567071003?mod=searchresults&page=1&pos=2.
Lombardo, C. and Ziobro, P. (2017). Trump Criticizes Post Office for Charging
Amazon ‘So Little.’ The Wall Street Journal, December 29,
https://www.wsj.com/articles/trump-criticizes-postal-service-for-
charging-amazon-so-little-
1514560466?mod=searchresults&page=1&pos=2.
MacArthur, J.B. (1992), Activity-Based Costing: How Many Cost Drivers Do
You Want? Journal of Cost Management, Fall, 37-41.
Office of Inspector General United States Postal Service (2013). Postal Service
Product Costing Methodologies Management Advisory Report. Report
Number MS-MA-13-002, Office of Inspector General United States
Postal Service, April 11,
https://www.uspsoig.gov/sites/default/files/document-library-
files/2015/ms-ma-13-002.pdf.
Perez, S. (2018). Amazon shipped over 5 billion items with Prime in 2017. The
Wall Street Journal, January 2,
https://techcrunch.com/2018/01/02/amazon-shipped-over-5-billion-
items-with-prime-in-2017/.
Peterson, H. (2018). Amazon is launching a new delivery program and hiring
thousands of drivers, with a warning against ‘peeing in bottles.’ Business
Insider, November 5, https://www.businessinsider.com/amazon-hires-
thousands-of-delivery-drivers-2018-11.
Postal Regulatory Commission (2016a). Order Concerning United Parcel
Service, Inc.’s Proposed Changes to Postal Service Costing
Methodologies (UPS Proposals One, Two, And Three). Order No. 3506,
Washington, DC: United States of America Postal Regulatory
Commission, September 9,
https://www.prc.gov/docs/97/97114/Order3506.pdf.
Postal Regulatory Commission (2016b). Notice of Proposed Rulemaking on
Changes Concerning Attributable Costing. Order No. 3507, Washington,
DC: United States of America Postal Regulatory Commission,
September 9,
https://www.prc.gov/docs/97/97115/Order%20No.%203507.pdf.
Journal of Business and Accounting
33
Postal Regulatory Commission (2016c). Order Adopting Final Rules on Changes
Concerning Attributable Costing. Order No. 3641, Washington, DC:
United States of America Postal Regulatory Commission, December 1,
https://www.prc.gov/docs/97/97990/Order3641.pdf.
Postal Regulatory Commission (2018). PRC Approves Rate Changes for Market
Dominant and Competitive Products. Press Release, Washington, DC:
United States of America Postal Regulatory Commission, November 13,
https://www.prc.gov/press-releases/prc-approves-rate-changes-market-
dominant-and-competitive-products/4536.
Postal Regulatory Commission (2019a). Order Adopting Final Rules Relating to
the Institutional Cost Contribution Requirement for Competitive
Products. Order No. 4963, Washington, DC: United States of America
Postal Regulatory Commission, January 3,
https://www.prc.gov/docs/107/107901/Order4963.pdf.
Postal Regulatory Commission (2019b). PRC Adopts Final Rules on Minimum
Contribution of Competitive Products to Institutional Cost: Formula-
based Approach Used to Calculate the Appropriate Share. Press Release,
Washington, DC: United States of America Postal Regulatory
Commission, January 3, https://www.prc.gov/press-releases/prc-adopts-
final-rules-minimum-contribution-competitive-products-institutional-
cost.
Sandbulte, J. (2017). Why the Post Office Gives Amazon Special Delivery. The
Wall Street Journal, July 13, https://www.wsj.com/articles/why-the-post-
office-gives-amazon-special-delivery-1499987531.
Save the Post Office (2018a). Edited and administered by Steve Hutkins. Talk
about Fake News: How a flawed Citigroup analysis led to Trump’s
bogus tweets about Amazon and the Postal Service. Save the Post Office,
April 3, https://savethepostoffice.com/talk-about-fake-news-how-a-flawed-
citigroup-analysis-led-to-trumps-bogus-tweets-about-amazon-and-the-postal-
service/.
Save the Post Office (2018b). Edited and administered by Steve Hutkins. Court
turns back UPS bid to drive up USPS parcel prices. Save the Post Office,
May 29, https://savethepostoffice.com/court-turns-back-ups-bid-to-drive-up-
usps-parcel-prices/.
Save the Post Office (2018c). Edited and administered by Steve Hutkins. UPS
goes back to court on USPS methodology. Save the Post Office, July 16,
https://savethepostoffice.com/ups-goes-back-to-court-on-usps-costing-
methodology/.
Save the Post Office (2018d). Edited and administered by Steve Hutkins. D.C.
Court rejects UPS request for a rehearing on USPS costing case. Save
the Post Office, August 8, https://savethepostoffice.com/dc-circuit-court-
rejects-ups-request-for-rehearing-usps-costing-case/.
Barton and MacArthur
34
Sink, J. and Soper, S. (2017). Trump Takes On Amazon Again, Urging ‘Much
More’ in Postage Fees. Bloomberg Politics, December 29,
https://www.bloomberg.com/news/articles/2017-12-29/trump-says-u-s-
post-office-should-charge-amazon-much-more.
Stevens, L. (2014). For FedEx and UPS, a Cheaper Route: the Post Office. The
Wall Street Journal, August 4, https://www.wsj.com/articles/u-s-mail-
does-the-trick-for-fedex-ups-1407182247.
Taylor, K. (2018a). Trump is delivering on his rants about the US Postal
Service's Amazon deal — but his argument is flawed. Business Insider,
April 13, https://www.businessinsider.com/trump-amazon-post-office-
deal-2018-4.
Taylor K. (2018b). Trump is doubling down on his claims that Amazon uses the
US Postal Service as a ‘delivery boy,’ and it could be a major blow to
countless American businesses. Business Insider, July 23,
https://www.businessinsider.com/trump-attacks-amazon-over-usps-deal-
threatens-business-2018-7.
Thomas, P. (2019). Amazon to Add 100,000 Electric Vehicles as Part of Climate
Pledge. The Wall Street Journal, September 19,
https://www.wsj.com/articles/amazon-to-add-100-000-electric-vehicles-
as-part-of-climate-pledge-
11568914649?mod=searchresults&page=3&pos=20.
United States Postal Service (2002). Activity-Based-Costing. 2002
Comprehensive Statement on Postal Operations, United States Postal
Service, Accessed February 18, 2019, https://about.usps.com/strategic-
planning/cs02/2p.htm.
United States Postal Service (2017). United States Postal Service FY 2017
Annual Report to Congress. United States Postal Service,
https://about.usps.com/who-we-are/financials/annual-reports/fy2017.pdf.
United States Postal Service (2018a). About: Postal Terms. United States Postal
Service, Accessed January 24, 2018,
http://about.usps.com/publications/pub32/pub32_terms.htm.
United States Postal Service (2018b). Postal Facts – U.S. Postal Service: The
Top 12 Things You Should Know About the U.S. Postal Service. United
States Postal Service, May 15, Accessed April 2, 2019,
https://facts.usps.com/top-facts/.
United States Postal Service (2018c). 2018 Report on Form 10-K. Washington
DC: United States Postal Regulatory Commission,
https://about.usps.com/who-we-are/financials/10k-reports/fy2018.pdf.
United States Postal Service (2019). The United States Postal Service Delivers
the facts. United States Postal Service, January, Accessed April 2, 2019,
https://about.usps.com/news/delivers-facts/usps-delivers-the-facts.pdf.
Ziobro, P. (2018). Postal Service Proposes Price Increase. The Wall Street
Journal, October 8, https://www.wsj.com/articles/postal-service-
proposes-price-increase-
1539208437?mod=searchresults&page=1&pos=1.
Journal of Business and Accounting
35
Ziobro, P. (2019). Postal Service Reports First Drop in Packages in Nearly a
Decade. The Wall Street Journal, August 9,
https://www.wsj.com/articles/postal-service-reports-first-drop-in-
packages-in-nearly-a-decade-
11565364114?mod=searchresults&page=1&pos=5.
Ziobro, P. and Mattioli, D. (2019). FedEx to End Ground Deliveries for Amazon.
The Wall Street Journal, August 7, https://www.wsj.com/articles/fedex-
to-end-u-s-ground-deliveries-for-amazon-
11565182535?mod=searchresults&page=1&pos=6.
Journal of Business and Accounting
Volume 12, No 1; Fall 2019
36
UNDERFUNDED PENSION LIABILITIES’ IMPACT ON
SECURITY RETURNS: THE U.S. VERSUS THE
EUROPEAN UNION
Ronald A. Stunda Valdosta State University
ABSTRACT
This study assesses the role that underfunded pensions have on security
returns. A twelve year analysis (2006-2017) is conducted which includes
analyzing 364 U.S. firms and 262 EU firms with underfunded pension plans.
Overall results indicate that, with respect to security prices, when firms are known
to have underfunded pension plans, European firms appear to be more cash flow
sensitive while U.S. firms appear to be more accounting earnings sensitive. These
findings may be the result of institutional and structural differences among nations.
On the other hand, it is noteworthy to understand that differences in cash flow
presentation and recognition exist between the nations and their standard-setters.
These differences have the potential of contributing to how investors make stock
purchase decisions and therefore must be considered as the transitional phase
between IFRS and GAAP continues.
Keywords: Pensions, security returns, IFRS, European Union
INTRODUCTION
In a corporate pension funding study released in 2018, it was found that
currently in the United States, publicly traded firms account for approximately $2.0
trillion in unfunded pension liabilities, [Clark, Perry and Wadia (2018)]. The fact
that, on average, the funding ratio for these firms is approximately 80% gives one
an idea of the amount of pension dollars committed by U.S. firms. European Union
(EU) firms face a somewhat similar dilemma. Even though a significant
conversion in the EU is on to switch from defined benefit to defined contribution
private pension plans, a significant number of defined benefits plans still exist.
Eurozone unfunded pension liabilities are projected to exceed those of the U.S. by
2020. What further complicates the picture for these European firms is that the
current average funding ratio is only 56% [Financial Times (2017)].
Since publicly traded firms, both in the U.S and the EU operate on the
premise of returning a profit to their shareholders, and thus contain a direct
correlation to the valuation of firms’ wealth through stock prices and the ability to
raise capital, this study will focus on the impact of underfunded pension liabilities
of publicly traded firms in the U.S. compared to their counterpart firms in the EU.
The underfunded pension obligations of telecommunications giants
AT&T and Verizon Communications, Inc. could double within the next five years,
Journal of Business and Accounting
37
thus dropping the plan below the mandated funded ratio of 80% and requiring even
larger cash contributions into the fund. In fact, these two firms have the highest
unfunded pension liabilities of any in the Standard and Poor’s 500 index [Global
Institutional Investors Survey (2015)]. U.S. Steel, also under pressure because of
their $2.5 billion unfunded pension liability, has recently begun to drop some
retirees from their pension plan. In the Eurozone, British Telecom currently has a
7 billion pound shortfall in its pension fund, while Daimler falls 2.5 billion euros
short, and Lufthansa approaches historically high levels of near 7 billion euros of
unfunded pension liabilities [Financial Times (2017)]. The direct impact of these
scenarios is to put pressure on firms’ bottom line through the realization of greater
pension expense during the current accounting period. It forces firms to pursue
strategies which permit minimizing expenses in other areas of operation in order
to compensate for the increase in pension expense. Many firms have had difficulty
in achieving this goal. The end result by some is lowered income, reduced stock
prices and eventual under capitalization as these firms find it harder to raise capital
for the future.
Although this problem seems to have been a long time in coming, very
little research has been conducted which fully assesses the impact of underfunded
pension plans on stock prices of U.S. firms versus their EU counterparts. This
study will attempt to do just that. A comparison analysis will be made of U.S. and
EU publicly traded firms that have underfunded pension plans in order to assess
any differences and impact on security prices over time.
LITERATURE REVIEW
Bulow, Morck and Summers (1987) utilize pension data from the 1980s
and find that market valuation of firms reasonably reflect their pension funding
structure. That is, availability and stability of a pension plan influences stock
prices upward while termination and instability of a pension plan tend to influence
stock prices downward. Other studies hint of stock price impact through the
influence of the debt market. Ederington (1992) finds that bond ratings are an
informational source to stockholders and thus impact stock prices. Iskandar and
Emery (1994) identify underlying influencers which affect bond rating. Pension
underfunding is one of those influencers. Carroll and Niehaus (1998) is the first
study to focus solely on years after the passage of Statement of Financial
Accounting Standard #87 (Accounting for Pensions) and finds that underfunded
pension plans reduce debt rating more than any other variable assessed. Thus, a
vital link is established from underfunded pension plans to stock prices through
the impact on a firm’s debt rating.
Chan, Jegadeesh and Lakonishok (2006) assess the quality of earnings and
its impact on stock prices. Their study shows that underfunded pension plans have
a negative impact on earnings quality and thus may affect stock prices. Fama and
French (1993) analyze risk factors in security returns and find that highly
underfunded pension liabilities represent a significant risk to the security returns
of a firm. Francesco (2009) assesses firms that underfund pension plans versus
Stunda
38
those that overfund pension plans. Although the time period is brief and the sample
small, their finding is that there is a significant difference in security price reaction.
Many European private pension plans have historically been based on a
pay as you go basis. However, with the inception of IAS 19 in 2004, the
accounting treatment of pension plans has been similar to what is prescribed in
SFAS 87. Mauldin (2017) finds that U.K. firms have the greatest shortfall in
pension obligations and this has had a significant impact on their capital formation
ability and subsequent stock prices. The Wall Street Journal notes that Spain and
Spanish firms were hit the hardest by the last financial crisis but bounced back
more vigorously than other countries such as Greece, Germany, Austria and France
with firms in these last four countries having the greatest impact in pension funding
shortfall. This has lead to an average of a 13% reduction of stock prices across
these countries [WSJ (2015)]. Kaier and Muller (2013) find a direct correlation to
underfunded pension liabilities and firm stock prices for firms across Europe.
METHODOLOGY
Sample and Data Collection
IASB Regulation Number 1606 (2002) requires companies listed on
European securities markets to begin using IFRS in their consolidated financial
statements starting in 2005. This includes accounting for employee benefits and
pensions. As a result, the sample used in this study will include the years 2006-
2017 (first full year subsequent to the year of adoption of IASB Regulation 1606
in the European markets to most current year available). Securities traded on the
European Stock Exchange (ESE), the London Stock Exchange (LSE) and the
Frankfurt Stock Exchange (FSE) and their associated prices are identified from the
AMADEUS database. Earnings and other corporate data related to these
firms/securities is found on Compustat Global. U.S. firm data is obtained from
Compustat for earnings information, and the Center for Research on Security
Prices (CRSP) for security price information.
Also, analysts’ forecast of earnings is obtained from the Investment
Brokers Estimate Service (IBES), and consists of quarterly point forecasts of
earnings for a given period. In addition, the Electronic Data Gathering and
Retrieval System (EDGAR), the Wall Street Journal, and the European Business
Register are used to analyze financial notes and other associated firm information
in order to control for such things as change of corporate form, change in
ownership, or change in management. If any of these could be documented during
the test period, the firm is subsequently eliminated from the study. Table 1
summarizes firms included in the sample.
Journal of Business and Accounting
39
Table 1
Summary of the Firms, 2006-2017
Item U.S. Firms European Firms
Firms initially identified 417 329
Firms eliminated due to
lack of CRSP/AMADEUS
data
27 31
Firms eliminated due to
lack of Compustat Global/
Compustat data
15 24
Firms eliminated due to
EDGAR/European
Business Register
information
11 12
Total firms in study sample 364 262
HYPOTHESIS DEVELOPMENT
Three hypotheses are tested. First, Ball and Brown (1968), Beaver,
Lambert and Morse (1980), Basu (1997) and Ball, Kothari and Robin (1998), all
incorporate a research design that associates accounting earnings to changes in
market values of equity. These studies indicate, (with varying degrees of
significance depending on such factors as firm size, firm risk or industry of the
firm), that there is an overall significant positive correlation between accounting
earnings and a firm’s security price. Drawing upon this literature, an in order to
establish a baseline upon which to further this line of research, the following
hypothesis is stated:
H1: There is no significant difference in the information content of
accounting earnings between U.S. firm firms and European sample
firms.
In their analysis of forecast accuracy of cash flows, Sinha, Brown, and Das
(1997) utilize a matched-pair design in which the forecast accuracy of the same
analyst is measured over time. Stunda (2016) uses the same methodology to
determine forecast accuracy of groups of analysts over a longer time period. Both
studies evaluate U.S. firms exclusively.
Although these prior studies assessed the analysts’ forecast accuracy of
cash flows of U.S. firms, this study will attempt to do the same with comparison
between U.S. and European firms. This is an attempt to assess the accuracy of
cash flow forecasts in relation to actual earnings, which ultimately provides an
indication of how close the earnings forecasts are of underfunded firms in the U.S.
versus those in European nations. These forecasts form a basis for future
Stunda
40
investment decisions by investors, which ultimately impact security prices. The
comparison gives rise to the second hypothesis, stated in null form:
H2: There is no significant difference in forecast accuracy of cash flows
between U.S. sample firms and European sample firms.
With regard to cash flow forecasts, DeFond and Hung (2003) report that
firms which have cash flow forecasts tend to have greater capital intensity.
McInnis and Collins (2008) provide evidence indicating that investors place
relatively more weight on the cash flows, as opposed to the accrual-based earnings.
As a result, a comparison of cash flows between samples is also undertaken. This
leads to the third hypothesis stated in the null form:
H3: There is no significant difference in the information content of cash
flows between U.S. sample firms and European sample firms.
TEST OF HYPOTHESES AND RESULTS
Test of Hypothesis 1
H1: Test of information content of accounting earnings Hypothesis 1 proposed that “There is no significant difference in the information
content of accounting earnings between U.S. firm firms and European sample.”
Using the Ball and Brown (1968) model to determine the earnings response
coefficient (ERC), the following model is established for determining information
content:
𝐶𝐴𝑅𝑖𝑡 = 𝑎 + 𝑏1𝑈𝐸𝑈𝑆𝑖𝑡 + 𝑏2𝑈𝐸𝐸𝑖𝑡 + 𝑏3𝑀𝐵𝑖𝑡 + 𝑏4𝐵𝑖𝑡 + 𝑏5𝑀𝑉𝑖𝑡 + 𝑒𝑖𝑡 (1)
Where: CARit = Cumulative abnormal return firm i, time t
a = Intercept term
UEUSit = Unexpected earnings for U.S. firms in the sample
UEEit = Unexpected earnings for European firms in the sample
MBit = Market to book value of equity as proxy for growth
and persistence
Bit = Market model slope coefficient as proxy for systematic
risk
MVit = Market value of equity as proxy for firm size
eit = Error term for firm i, time t
The coefficient “a” measures the intercept. The coefficient b1 represents
U.S. firms and is the traditional earnings response coefficient (ERC), found to have
correlation with security prices in market-based studies (Ball and Brown, 1968).
The coefficient b2 is the ERC associated with European firms. Unexpected
Journal of Business and Accounting
41
earnings (UEi) is measured as the difference between the management earnings
forecast (MFi) and security market participants’ expectations for earnings as a
proxy for consensus analyst following as per Investment Brokers Estimate Service
(IBES) (EXi). The unexpected earnings are scaled by the firm’s stock price (Pi)
180 days prior to the forecast:
𝑈𝐸𝑖 = [(𝑀𝐹𝑖) − (𝐸𝑋𝑖)]/𝑃𝑖 (2)
Unexpected earnings are measured for each of the sample firms during the
test period. The coefficients b3, b4, and b5, are contributions to the ERC for all
firms in the sample. To investigate the effects of the information content of
earnings on security returns, there must be some control for variables shown by
prior studies to be determinants of ERC. For this reason, the variables represented
by coefficients b3 through b5 are included in the study.
For each firm sample, an abnormal return (ARit) is generated around the
event dates of -1, 0, +1 (day 0 representing the day that the firm’s financials were
available per DJNRS or Worldscope). The market model is utilized along with the
CRSP equally-weighted market index and regression parameters are established
between -290 and -91. Abnormal returns are then summed to calculate a cross-
sectional cumulative abnormal return (CARit).
Results H1 Results of correlating the ERC to security returns are presented in Table
2. Results show that coefficient b1, representing the ERC of U.S. firms, is 1.96 and
significant at the .01 level. These results provide a fairly strong relationship
between accounting earnings and security prices, indicating that investors perceive
that accounting earnings possess information content and therefore have predictive
value when correlated with security returns. Coefficient b2, representing the ERC
of European firms, is 0.22 and significant at the .10 level. Although, for European
firms, the relationship between accounting earnings and security prices is positive,
it is not very strong. This indicates that accounting earnings carry a lesser extent
of information content among European investors, and therefore they are perceived
to have less predictive value. Ali and Hwang (2000) find that the European model
emphasizes tax rules and accounting standards and as such, more emphasis is spent
adhering to rules and standards and less emphasis to earnings disclosures. This
could account for some of the difference seen in this test.
Performing the Welch’s test, and as indicated in Table 2, a t-statistic of
1.619 was computed with a p-value of less than .010. This indicates that the means
of the sample groups are significantly different, and thus the null of similarity
between the groups is rejected.
Stunda
42
Table 2
Information Content of Accounting Earnings 2006-2017
U.S. Firms versus European Firms
Model: 𝐶𝐴𝑅𝑖𝑡 = 𝑎 + 𝑏1𝑈𝐸𝑈𝑆𝑖𝑡 + 𝑏2𝑈𝐸𝐸𝑖𝑡 + 𝑏3𝑀𝐵𝑖𝑡 + 𝑏4𝐵𝑖𝑡 +𝑏5𝑀𝑉𝑖𝑡 + 𝑒𝑖𝑡
a b1 b2 b3 b4 b5 Adj. R2
0.20 1.96 0.22 0.42 0.38 0.12 0.241
(.87) 1.66a 2.21b 0.30 0.51 0.62
a = significant at the .01 level
b = significant at the .10 level
t-stat df p-value
Welch’s t-test 1.619 1 <.010
In addition, whenever regression variables are employed, there is a
probability of the presence of multicollinearity within the set of independent
variables which may be problematic from an interpretive perspective. To assess
the presence of multicollinearity, the Variance Inflation Factor (VIF) was utilized.
Values of VIF exceeding 10 are often regarded as indicating multicollinearity. In
the test of hypothesis 2, a VIF of 2.8 was observed, thus indicating a non-presence
of significant multicollinearity
Test of Hypothesis 2
H2: Test of forecast accuracy of cash flows
Hypothesis 2 proposed that “There is no significant difference in forecast
accuracy of cash flows between U.S. sample firms and European sample firms.”
Consistent with the methodology of Sinha, Brown, and Das (1997), the following
model is used to assess forecast accuracy among analysts:
rapfeijt = |Rjt – Fmjt)/Rjt|*100 - |Rjt – Fijt)/Rjt|*100
(3)
Where: subscripts i, j, t denote analyst, firm and year, respectively
Rjt is the j firm’s cash flows in year t
Fijt is the forecast of cash flow by analyst i for firm j in year t
Fmjt is forecast of average analyst for the firm
rapfeijt is each analyst’s relative absolute percentage forecast
error, calculated as the absolute percentage
forecast error of the average analyst minus that
of the above average analyst. For below average
Journal of Business and Accounting
43
analysts, the order of the two terms on the right hand
side are reversed (i.e. in the later case the individual
analyst’s projection is below the average analyst’s
projection).
Results H2 A pooled, cross-sectional analysis is performed over the study period
2016-2017 and incorporating all analyst forecasts. Data analysis in Table 3
indicates the results of the analysis. Results indicate that U.S. analysts have a larger
average forecast error (3.05) which is significant at the 0.01 level. European
analysts have a smaller average forecast error (1.10) which is significant at the 0.01
level.
Because the variances of the groups are not equal, there exists violation of
the assumption of homogeneity across the sample. In order to account for this, the
Welch’s test was performed. This test assesses the significance between groups
when variances do not equal. Based on the Welch’s test, and as indicated in Table
3, a t-statistic of 1.682 was computed with a p-value of less than .010. This
indicates that the means of the sample groups are significantly different, and thus
the null of similarity between the groups is rejected.
Muller and Verschoor (2005) find that European firms have a history of
placing importance on cash flows along with their associated forecast. Obrien
(1990) surmises that the relatively more experienced analysts have greater forecast
accuracy. The above results could be a combination of European firms placing
greater emphasis on cash flow forecasts, while possessing analysts who are more
experienced in these forecasts.
Table 3
Analysis Accuracy of Cash Flows 2006-2017
U.S. Firms versus European Firms
Model: rapfeijt= |Rjt – Fmjt)/Rjt|*100 - |Rjt – Fijt)/Rjt|*100
Entity Number Average
Mean
t-test Probability
U.S. Firms 364 3.05 1.70 0.01
European
Firms
262 1.10 1.67 0.01
t-stat df p-value
Welch’s t-test 1.682 1 <.010
Test of Hypothesis 3
H3: Test of information content of cash flows
Hypothesis 3 proposed that “There is no significant difference in the
information content of cash flows between U.S. sample firms and European sample
Stunda
44
firms”. Using a similar approach as hypothesis one, a regression model is
constructed to assess the information content of cash flows. This model is
consistent with that used by Clement (1999), Jacob et al. (1999), Chen and
Matsumoto (2006), and Call, Chen and Tong (2009). The model is:
𝐶𝐴𝑅𝑖𝑡 = 𝑎 + 𝑏1𝑈𝐸𝑈𝑆𝑖𝑡 + 𝑏2𝑈𝐸𝐸𝑖𝑡 + 𝑏3𝑀𝐵𝑖𝑡 + 𝑏4𝐵𝑖𝑡 + 𝑏5𝑀𝑉𝑖𝑡 + 𝑒𝑖𝑡 (4)
Where: CARit = Cumulative abnormal return firm i, time t
a = Intercept term
UEUSit = Cash for U.S. firms in the sample
UEEit = Cash for European firms in the sample
MBit = Market to book value of equity as proxy for growth
and persistence
Bit = Market model slope coefficient as proxy for systematic
risk
MVit = Market value of equity as proxy for firm size
eit = Error term for firm i, time t
The coefficient “a” measures the intercept. The coefficient b1 represents
U.S. firms and is the response coefficient correlating cash flows with security
prices. The coefficient b2 is the similar response coefficient associated with
European firms. Once again, variables b3, b4, and b5 are inserted for assessment
of other determinants which may influence results.
Similarly, for each firm sample, an abnormal return (ARit) is generated
around the event dates of -1, 0, +1 (day 0 representing the day that the firm’s
financials were available per DJNRS or Worldscope). The market model is utilized
along with the CRSP equally-weighted market index and regression parameters
are established between -290 and -91. Abnormal returns are then summed to
calculate a cross-sectional cumulative abnormal return (CARit).
Results H3 Results of the test for hypothesis 3 are found in Table 4. Results show that
coefficient b1, representing the response coefficient of U.S. firms, is 0.62 and
significant at the .10 level. These results indicate that a positive relationship exists
between cash flows and security prices for U.S. firms. Coefficient b2, representing
the response coefficient of European firms, is 3.23 and significant at the .01 level.
This indicates a stronger relationship between the information content contained
in reported cash flows and security prices for European firms.
Performing the Welch’s test, and as indicated in Table 4, a t-statistic of
1.642 was computed with a p-value of less than .010. This shows that the means
of the sample groups are significantly different, and thus the null of similarity
between the groups is rejected. This reinforces the finding of Marshall (2000)
which indicates that European Union firms tends to place significant reliance on
cash flows.
Journal of Business and Accounting
45
Table 4
Information Content of Cash Flow 2006-2016
U.S. Firms versus European Firms
Model: 𝐶𝐴𝑅𝑖𝑡 = 𝑎 + 𝑏1𝑈𝐸𝑈𝑆𝑖𝑡 + 𝑏2𝑈𝐸𝐸𝑖𝑡 + 𝑏3𝑀𝐵𝑖𝑡 + 𝑏4𝐵𝑖𝑡 +𝑏5𝑀𝑉𝑖𝑡 + 𝑒𝑖𝑡
a b1 b2 b3 b4 b5 Adj. R2
0.40 0.62 3.23 0.55 0.42 0.46 0.238
(.38) 2.19b 1.60a 0.88 0.39 0.70
a = significant at the .01 level
b = significant at the .10 level
t-stat df p-value
Welch’s t-test 1.642 1 <.010
Again, regression variables are assessed for multicollinearity using the
Variance Inflation Factor (VIF). Values of VIF exceeding 10 are often regarded
as indicating multicollinearity. In the test of hypothesis 3, a VIF of 2.4 was
observed, thus indicating a non-presence of significant multicollinearity
CONCLUSIONS
The purpose of this study was to compare U.S. versus European firms that
have underfunded pension plans, and assess differences that they may display in
relationship to security prices. The study compared firms with underfunded
pension plans during the study periods 2006-2017. A sample of 364 U.S. firms
and 262 European firms were contained in the study. In evaluating pension plans,
this study is one of the first to use a study period greater than 5 years (i.e., 12), and
it is also one of the first to analyze a significant number of firms in both the U.S.
and Europe. Previous studies include 50 or fewer firms. The results, therefore,
should be robust and indicative of publicly traded companies with underfunded
pension plans on both continents. The results indicate that there is indeed a
significant difference between firms with underfunded pension plans in the U.S.
versus Europe.
First, the information content of accounting earnings is assessed for the
two sample groups. The analysis finds that a higher degree of correlation between
accounting earnings and stock price exists for U.S. firms than for European firms.
This may be due in part to prior studies which show that the European model
emphasizes tax rules and accounting standards and as such, more emphasis is spent
adhering to rules and standards and less emphasis to earnings disclosures.
Next, forecast accuracy of cash flows between the two groups is assessed.
Findings indicate that there is a higher degree of forecast accuracy of cash flows
among European firms. This may be due to findings of prior studies that show
Stunda
46
European firms have a history of placing importance on cash flows along with their
associated forecast.
Finally, the information content of cash flows is assessed for the two
sample groups. Findings show that a higher degree of correlation between cash
flow and security returns exists for European firms than for U.S. firms. This may
be again related to prior studies which indicate that European Union firms tends to
place significant reliance on cash flows.
Overall results indicate that, with respect to security prices, when firms are
known to have underfunded pension plans, European firms appear to be more cash
flow sensitive while U.S. firms appear to be more accounting earnings sensitive.
These findings may be the result of institutional and structural differences among
nations. On the other hand, it is noteworthy to understand that differences in cash
flow presentation and recognition exist between the nations and their standard-
setters. These differences have the potential of contributing to how investors make
stock purchase decisions and therefore must be considered as the transitional phase
between IFRS and GAAP continues.
REFERENCES
Ali, A. and L. Hwang (2000). Country specific factors related to financial
reporting and the value relevance of accounting data. Journal of
Accounting Research, 38, 1–21.
Ball, R., and P. Brown (1968). An empirical evaluation of accounting
income numbers, Journal of Accounting Research 6, 159-178.
Ball, R., S. Kothari, and A. Robin (1998). The effect of international
institutional factors on properties of accounting earnings, Rochester
Institute of Technology, 1-31.
Basu, S. (1997). The conservatism principle and the asymmetric timeliness
of earnings, Journal of Accounting and Economics 24, 3-37.
Beaver, W., R. Lambert, and D. Morse (1980). The information content of
security prices, Journal of Accounting and Economics 2, 3-28.
Bulow, J., R. Morck, and L. Summers (1987). How does the market value affect
unfunded pension liabilities? Journal of Economics 97(August), 435-452.
Call, A., S. Chen, and Y. Tong (2009). Are analysts’ earnings forecasts more
accurate when accompanied by cash flow forecasts? Revenue Accounting
Study, 14, 358-391.
Carroll, T. and G. Neihaus (1998). Pension plan funding and corporate debt rating.
The Journal of Risk and Insurance 65 (March), 427-441.
Journal of Business and Accounting
47
Chan, K., N. Jegadeesh, and J. Lakonishok ((2006). Earnings quality and stock
returns. The Journal of Business 79(3), 1041-1082.
Chen,S. and D. Matsumoto (2006). Favorable versus unfavorable
recommendations: The impact on analyst access to management-provided
information. Journal of Accounting Research, 44, 657–689.
Clark, C., A. Perry, and Z Wadia (2018). Corporate pension funding. Millman
Press, October, 1-28
Clement, M. (1999). Analyst forecast accuracy. Journal of Accounting and
Economics 27, 285-303.
Defong, M., and M. Hung (2003). Am empirical analysis of analysts’ cash flow
forecasts, Journal of Accounting and Economics, 35, 73-100.
Ederington, L. (1992). The importance of bond ratings. Journal of Risk and
Insurance 55(March), 32-58.
Fama, E., and K. French (1993). Common risk factors in the returns on stocks and
bonds. Journal of Financial Economics 33(February), 3-56.
Financial Times (2017). London Press, 1-10.
Francesco, F. (2009). Evidence from prices reactions to pension contributions.
Journal of Financial Economics 92(May), 491-518.
Global Institutional Investors Survey. October, 2015. AT&T, Verizon pension
funding issues to hurt bottom lines. Pages 7-14.
Iskandar, M. and E. Emery (1994). An empirical investigation of the role of
indenture provisions in determining debt ratings. Journal of Banking and
Finance 18(January), 93-112.
Kaier, K, and C. Muller (2013). New findings on unfunded pension plans across
Europe. Freiburg University Press, No. 52, 1-42.
Mauldin, J. (2017). The pension storm is coming to Europe. Vanguard, October
3, 11-18
McInnis, J. and D. Collins (2008). Do cash flow forecasts deter earnings
management? Working paper. University of Texas at Austin and
University of Iowa.
Stunda
48
Muller, J. and W. Verschoor (2005). The Impact of Corporate Derivative Usage
on Foreign Exchange Risk Exposure, European Financial Management,
48:261-297.
Obrien, P. (1990). “Forecast Accuracy of Individual Analysts in Nine Industries,”
Journal of Accounting Research, 286-304.
Sinha, P, L. Brown, and S. Das (1997). “A Re-Examination of Financial Analysts’
Differential Earnings Forecast Accuracy,” Contemporary Accounting
Research, 30-44.
Stunda, R. (2016). Financial analyst accuracy: an examination by industry. Global
Journal of Management and Business Research, (16), 2, 1-8
Journal of Business and Accounting
Volume 12, No 1; Fall 2019
49
TECHNOSTRESS: AN ANTECEDENT OF JOB
TURNOVER INTENTION IN THE ACCOUNTING
PROFESSION
Stacy Boyer-Davis
Northern Michigan University
ABSTRACT: Transformative technologies are rapidly changing the
accounting profession. The technological demand placed upon accounting
professionals is unprecedented. Moreover, the United States accounting
profession is plagued with significantly higher than average job turnover as
compared to other industries. A modern form of stress has emerged in the
workplace. Known as technostress, this type of stress originates from the
interaction with information and communication technologies. The adverse
physiological and psychosomatic consequences that result from technostress are
argued to influence job turnover intention similarly to other role stressors. The
purpose of this study is to examine the effect that technology stress, along with
organizational commitment, job satisfaction, and satisfaction with life, have on the
job turnover intention in the accounting profession. This paper broadens the
literature with respect to job turnover by evaluating the potential of technostress as
an antecedent.
Key Words: Job turnover, Turnover intention, Technostress,
Organizational commitment, Job satisfaction, Satisfaction with life, Accounting
INTRODUCTION
The professional business services industry, which includes accounting,
auditing, attestation, tax preparation, bookkeeping, consulting, and payroll
processing services, continues to be plagued with significantly higher than average
job turnover rates in the United States. The job turnover rate for the professional
services industry is nearly 2.3 times greater than the national average (U.S. Bureau
of Labor Statistics, 2018). Linkedin recently conducted a global job turnover study
of 500 million professionals across all industries. Worldwide data indicated job
turnover rates for the professional services sector were 1.1 times greater than the
combined average for all industries (Linkedin, 2018).
The cost of turnover per employee can equal or exceed their annual
compensation (Mitrovska & Eftimov, 2016). However, for management and
highly skilled positions like those in the accounting field, this amount can reach in
upwards of 250% of yearly earnings (SHRM, 2014). The typical price tag of
turnover includes separation, recruitment, and selection costs, training and
employee development, and reduced efficiency. From the perspective of
productivity, until a new employee is fully functional, the operation of an entire
organization can be impaired (Huselid, 1995).
Boyer-Davis
50
The hidden costs of employee separation are even more costly. Lost
expertise can be disruptive to a company, affecting workflow and customer service
quality. Turnover can degrade employee morale, giving rise to increased
workplace stress and absenteeism (Mitra, Jenkins, & Gupta, 1992). Moreover, the
risk of additional turnover magnifies (Felps et al., 2009).
Due to the disproportionate level of turnover in the accounting profession
and the costs associated with it, an extensive amount of research has been amassed
to study the antecedents of job turnover intentions (Bao, Bao, & Vasarhelyi, 1986;
Chong & Monroe, 2015; Doll, 1983; Harrell, 1990; Nouri, 2017; Parker &
Kohlmeyer, 2005; Pitman, Gaertner, & Hemmeter, 2011). Prior studies have
concluded that organizational commitment and job satisfaction are adversely
related to job turnover intention (Gregson, 1992; Lee & Jeong, 2017; Pasewark &
Strawser, 1996; Viator, 2001). Work-life balance and life satisfaction have also
been linked to job turnover (Schieman, Milkie, & Glavin, 2009). Role stressors,
including role ambiguity, role conflict, and role overload, operate as facilitators of
job turnover (Bedeian & Armenakis, 1981; Fogarty, Singh, Rhoads, & Moore,
2000; Kim, Im, & Hwang, 2015; Senatra, 1980).
A modern form of stress has emerged in the workplace. Known as
technostress, this type of stress originates from the interaction with information
and communication technologies (ICTs) (Agervold, 1987; Brod, 1984). The
adverse physiological and psychosomatic consequences that result from
technostress are argued to influence job turnover intention similarly to other role
stressors (Brillhart, 2004; Tarafdar, Tu, Ragu-Nathan, & Ragu-Nathan, 2007). The
purpose of this study is to examine the effect that technology stress has on the job
turnover intention in the accounting profession. This paper broadens the literature
with respect to job turnover by evaluating the potential of technostress as an
antecedent.
LITERATURE REVIEW
Transformative technologies are rapidly changing the accounting
profession. Repetitive tasks are being automated by artificial intelligence, machine
learning, blockchain, and robotics. Internet-based cloud computing and mobile
applications provide instant access to business or client information. Advances in
software enable efficient analysis of big data, triple-entry bookkeeping, and more
streamlined tax preparation. Cybersecurity threats and unauthorized data hacks
are common realities. The technological demand placed upon accounting
professionals is unprecedented (Drew, 2018).
Accountants must be retrained with a new set of skills to navigate the
automation revolution or face replacement (Pew Research Center, 2016).
Uncertainty can arise from adapting to new workplace technologies as can
insecurity from the threat of job loss. Having to work faster and longer to keep up
with current job responsibilities while simultaneously retraining for those expected
in the future can be an overwhelming process. The physical and emotional
responses elicited by the use of ICTs is a phenomenon categorized as technostress.
Journal of Business and Accounting
51
Technostress
Technostress, also known as technophobia, technology stress, computer
anxiety, computerphobia, digital depression, and computer stress, is an anxiety
disorder that results from an inability to adapt to or cope with technology (Brod,
1984; Chua, Chen, & Wong, 2009; Durndell & Haag, 2002). Those afflicted with
technostress may be impatient, moody, irritable, anxious, fatigued, exhausted,
confused, unable to concentrate, impulsive, pessimistic, or even depressed (Riedl,
Kingermann, Auinger, & Javor, 2012). Other health outcomes associated with
technostress include headaches, indigestion, hypertension, and cardiac arrest
(Pucci, Cristina, Antonaci, Costa, Imbriani, & Taino, 2015).
Tarafdar et al. (2007) constructed a framework of five technostress
creators, triggered by interfacing with ICTs. Users feel compelled to: (a) increase
work pace and job load (techno-overload), (b) stay connected, upsetting the work-
life balance (techno-invasion), (c) spend more time learning about complex
technologies due to feelings of inadequacy (techno-complexity), (d) regard their
jobs are in jeopardy (techno-insecurity), and (e) are uncomfortable with continuous
change (techno-uncertainty). These techno-stressors can lead to role overload, the
perception that a job is too difficult or demanding (Tarafdar, Tu, Ragu-Nathan, &
Ragu-Nathan, 2011). Technostress can also produce role conflict, whereby the
pressures imposed by ICTs at work interrupt time spent at home (Kahn et al., 1964;
Tarafdar et al., 2011). Finally, technostress may impose role ambiguity as job
expectations may become blurred when new technologies are adopted (Kahn et al.,
1964; Lee, Shin, & Baek, 2017).
If role stressors induce job turnover intention and technostress
creates role stress, then technostress may also prompt job turnover intention
(Bedeian & Armenakis, 1981; Fogarty, Singh, Rhoads, & Moore, 2000; Kim, Im,
& Hwang, 2015; Senatra, 1980; Tarafdar et al., 2007, 2011).
Hypothesis 1 (H1): Technostress is positively associated with job turnover
intention in the accounting profession.
Organizational Commitment
Organizational commitment has been described as the degree of
attachment one has to their workplace (Greenberg, 2005; Mowday, Steers, &
Porter, 1979). This emotional connection to an organization can improve
productivity, enhance job involvement and loyalty, and reduce work stress
(O’Reilly, 1989). Employees who are committed to an organization can more
deeply identify with corporate values, goals, and objectives, better equipping them
to embrace change (Vakola & Nikolaou, 2005). Research results have been
consistent across the globe (Siu, 2003).
Weakened or compromised organizational commitment can increase the
incidence of job turnover (Allen & Meyer, 1996; Mathieu & Zajac, 1990). Work
stress can exacerbate a degradation of organizational commitment (Boshoff &
Mels, 1994; Dale & Fox, 2008; Lee & Jamil, 2003; Tu, Ragu-Nathan, & Ragu-
Nathan, 2001). As technology related tensions rise at the workplace, the
Boyer-Davis
52
organizational commitment of accountants may wane. Therefore, in creating a
model to predict job turnover intention in the accounting profession, organizational
commitment is considered.
Hypothesis 2 (H2): Organizational commitment is negatively associated
with job turnover intention in the accounting profession.
Job Satisfaction
Job satisfaction has been characterized as the extent of
contentedness at the workplace (Locke, 1976; Spector, 1985). Various
determinants of job satisfaction include financial and nonfinancial benefits,
personal growth, supervisors and coworkers, working conditions, communication,
recognition and promotion, and job security (Benz & Frey, 2008). Employees who
are highly satisfied at work are less likely to quit their jobs (Chen et al., 2011). In
contrast, those workers who display negative feelings and are less satisfied are
more prone to separate from the company (Tett & Meyer, 1993).
ICTs will continue to shape not only the role of the accountant but
also the profession, as a whole. As much of the data entry work of the profession
will be eliminated, accountants must transform themselves into strategic
consultants and advisers. This metamorphosis may be disconcerting to many as
workers step out of their comfort zones and away from what is familiar. A
perception of career instability or fear of job insecurity may emerge, leading to a
reduction of job satisfaction.
Hypothesis 3 (H3): Job satisfaction is negatively associated with job
turnover intention in the accounting profession.
Satisfaction with Life
ICTs offer immense benefits to both employers and employees. From an
employer perspective, technology can improve workplace efficiency and
effectiveness while reducing costs. Employees may be provided with greater
workplace flexibility. Many organizations now endorse telecommuting, virtual
workplaces, and other flexible work arrangements to manage their workloads and
life commitments. A healthy work-life balance can be achieved when time is
appropriately allocated to both professional endeavors and personal interests.
Work-life balance promotes satisfaction with life, in general, and supports overall
well-being (Grzywacz, Butler, & Almeida, 2009).
The evolution of ICTs has clouded the boundaries between work and life
(Greenhaus, Collins, & Shaw, 2003). Employees may feel pressured to tether
themselves virtually to the workplace after hours. This compulsion to immediately
respond to workplace requests can upset the work-life balance. The further that
work-life balance moves away from equilibrium, the greater the threat of job
turnover (Asiedu-Appiah, Mehmood, & Bamfo, 2015; Wright et al., 2014).
Journal of Business and Accounting
53
The satisfaction with life scale measures the quality and fulfillment with
life as a whole (Diener et al., 1985). People who are satisfied with their lives tend
to be more positive and healthier (Toker, 2012). Life satisfaction is inversely
related to job turnover intention (Lambert et al., 2009). Based on previous research
examining life satisfaction and job turnover intention, the following hypotheses
was posed:
Hypothesis 4 (H4): Satisfaction with life is negatively associated with job
turnover intention in the accounting profession.
RESEARCH DESIGN
To facilitate an understanding of how technostress, organizational
commitment, job satisfaction, and satisfaction with life influences job turnover
intention within the accounting profession, the following research question was
answered:
Research Question (RQ): What effects do technostress, organizational
commitment, job satisfaction, and life satisfaction have on the job turnover
intention of accounting professionals?
Data was collected via a multiple-choice, Likert-scale survey combining
questions from the Technostress, Job Satisfaction Survey, Organizational
Commitment Questionnaire, Satisfaction with Life Scale, and Job Turnover
Intention instruments (Diener et al., 1985; Mowday, Steers, & Porter, 1979;
Spector, 1985; Tarafdar et al., 2007; Walsh, Ashford, & Hill, 1985). A survey panel
of accounting professionals was used. The survey was electronically distributed
to a random sample selected from the panel. Responses were assumed to be and
converted to an interval level of measure in order to apply parametric tests during
data analysis. A multiple regression model was developed to evaluate the
significance of the relationships between the variables.
A minimum sample size of 184 was estimated using G*Power 3.1.9.2,
assuming a priori power analysis, α = .05, β = .95, and a medium effect size (Faul,
Erdfelder, Lang, & Buchner, 2007). An instrument is expected to demonstrate a
reliability of α = .70 or greater (Babbie, 2010). Table 1 provides the average
reliability for the instruments used as part of this study. A statistically large sample
was acquired to minimize threats to internal and external validity.
Boyer-Davis
54
SAMPLE DESCRIPTION
Respondents self-reported as 55.3% female, 44.2% male, and .5% trans-
male with a majority aged between 31-35 (23.4%), 36-40 (16.8%), or 41-45
(11.2%) years old. Of those surveyed, 47.7% earned a bachelor’s degree, 20.3% a
master’s degree, and 15.7% an associate degree. Approximately 51% of the
sample held a management position at the workplace. Regularly observed job
positions held by the survey participants included financial accountants (25.9%),
accounting directors (20.8%), staff accountants (14.2%), financial advisors (8.6%),
and auditors (4.6%). As per Table 2, frequently reported experience levels ranged
between 6 to 10 years (30.1%), 1 to 5 years (29.6%), and 11 to 15 years (15.3%).
Table 1
Reliability of Survey Instruments
Technostress Instrument .71α to .91α
Tarafdar et al. (2007)
Organizational Commitment Questionnaire .84α to .89α
Mowday, Steers, & Porter (1979)
Job Satisfaction Survey .82α to .93α
Spector (1985)
Satisfaction with Life Scale .74α to .86α
Diener et al. (1985)
Job Turnover Intention .70α to .80α
Walsh, Ashford, & Hill (1985)
Instrument Reliability
Table 2
Professional Experience
Frequencies and Percentages (N = 196 )
Years Frequency Percentage
< 1 2 1.0%
1 to 5 58 29.6%
6 to 10 59 30.1%
11 to 15 30 15.3%
16 to 20 14 7.1%
21 to 25 13 6.6%
26 to 30 12 6.1%
31 to 35 6 3.1%
36 to 40 0 0.0%
> 40 2 1.0%
Journal of Business and Accounting
55
Those accounting and financial professionals surveyed worked in the
private and public sectors, 60% and 22%, respectively, with approximately 9%
employed by governmental entities and 9% for not-for-profit companies (see Table
3).
Most of the accounting and financial professional surveyed work either 36
to 40 (33.2%), 41 to 45 (23.0%), or 46 to 50 (19.9%) hours per week (see Table 4).
DATA ANALYSIS
SPSS was the software used for data analytics for this study. To manage
missing data abnormalities, a listwise deletion approach was applied. Nominal
Table 3
Company Size and Business Sector
Frequencies and Percentages (N = 196 )
Size Public Private Government Not for Profit
1 to 50 1 36 1 4
51 to 100 6 21 2 3
101 to 250 3 17 1 1
251 to 500 4 8 2 1
501 to 1,000 5 14 2 1
1,001 to 5,000 10 12 1 3
> 5,000 14 10 8 5
Table 4
Average Number of Weekly Work Hours
Frequencies and Percentages (N = 196 )
Years Frequency Percentage
0 to 10 3 1.5%
11 to 15 2 1.0%
16 to 20 4 2.0%
21 to 25 2 1.0%
26 to 30 6 3.1%
31 to 35 5 2.6%
36 to 40 65 33.2%
41 to 45 45 23.0%
46 to 50 39 19.9%
51 to 55 9 4.6%
56 to 60 10 5.1%
60 to 65 3 1.5%
66 to 70 1 0.5%
> 70 2 1.0%
Boyer-Davis
56
variables were coded with dummy variables prior to conducting the analysis.
Variables were standardized into normalized z-scores to identify univariate
outliers. Outliers were analyzed using residuals and Cook’s Distance (Howell,
2007). Normality was evaluated with a normal probability plot, linearity was
confirmed through visual inspection of a histogram, and homoscedasticity was
assessed through the review of a residuals scatterplot (Field, 2013). A Durbin-
Watson test statistic for the model of 1.78 indicated uncorrelated adjacent residuals
and collinearity statistics were within the acceptable range of .1 and 10
(Tabachnick & Fidell, 2013).
RESULTS
Linear multiple regression was conducted to test the research question and
hypotheses. The regression model summary is shown in Table 5. According to
the model, 51% of the variance in job turnover intention among accounting
professionals was explained by the independent variables (technostress,
organizational commitment, satisfaction with life, and job satisfaction),
incorporated as part of this study.
Table 6 indicates that the ANOVA was statistically significant, F(196) =
48.90 at p < .01, for technostress, organizational commitment, and job satisfaction
predicting job turnover intention.
Table 7 reports the coefficient data for the model.
Table 5
Model Summary (N = 196 )
F df1 df2 Sig. F
Durbin-
Watson
0.71 0.51 0.50 5.37 48.90 4 191 0.00 1.78
Note. Dependent variable = Job Turnover Intention.
R R2
Adjusted
R2
Std. Error
of the
Estimate
Table 6
Analysis of Variance of the Regression Model (N = 196 )
Sum of
Squares
df Mean
Square
F Sig.
5633.70 4 1408.42 48.90 .00
5501.75 191 28.81
Total 11135.44 195
Model
Note. Dependent variable = Job Turnover Intention.
Regression
Residual
Journal of Business and Accounting
57
Technostress (β = .21, t(196) = p < .00), organizational commitment (β =
-.45, t(196) = p < .00), and job satisfaction (β = -.19, t(196) = p < .03), were
statistically significant in predicting job turnover intention in accounting
professionals. However, satisfaction with life was not significant (β = -.06, t(196)
= p < .28).
Table 8 summarizes the regression results and the overall findings of the
study based on the research question and sub-hypotheses.
DISCUSSION
This study identified that a relationship exists between
technostress and job turnover intention in the accounting profession. From a
practitioner standpoint, managers now have an awareness that technostress is of
considerable concern and can lead to an increased incident of job turnover rates in
the accounting profession. Firms are now able to strategize how to minimize the
stressors associated with technology.
One way that businesses can potentially curb technology stress is
to provide staff training prior to an implementation and system support thereafter.
Knowledge acquisition not only can reduce the uncertainty that employees feel
about the technology but can moderate the perception of its complexity. Likewise,
effective workplace communication is pivotal to lessen insecurities and
ambiguities surrounding a technology event. Companies should staff
appropriately so that role overload can be eased during the transition.
Table 7
Variables B SE β t Sig.
Constant 23.99 3.42 7.02 0.00
Technostress 0.06 0.02 0.21 3.79 0.00
Organizational Commitment -0.20 0.04 -0.45 -5.76 0.00
Satisfaction with Life -0.07 0.06 -0.06 -1.09 0.28
Job Satisfaction -0.04 0.02 -0.19 -2.17 0.03
Note. Constant (y-intercept). Overall model R2 = .71, F (4, 191) = 48.90, p = .001.
Multiple Regression Model Coefficient Data (N = 196 )
Table 8
Summary of Hypothesis Testing Results
Hypothesis Variable t Value Results
H1 Technostress .00 Rejected
H2 Organizational Commitment .00 Rejected
H3 Satisfaction with Life .28 Not Rejected
H4 Job Satisfaction .03 Rejected
Boyer-Davis
58
Employers should beware that policies that promote constant
connectivity by their employees may drive up technostress and cause job turnover.
Managers should respect employee personal time as much as possible and avoid
contact after hours. Furthermore, leaders should set the tone by unplugging
themselves.
In addition to the impact of job turnover, the economic blow of
technostress can be massive, equating to 225 million lost workdays and $300
billion in lost productivity per year (American Institute of Stress, 2007). Other
destructive outcomes include decreased profit and company value, a degradation
of services, and an increase in job vacancies (Moses, 2013). Technostress can
create a toxic work environment, accelerating workplace conflicts, permanently
damaging the corporate culture. This study has reinforced the merit of developing
approaches to lessen or eliminate the effects of technostress.
Companies should focus on improving the organizational commitment and
job satisfaction of their personnel since both were found to be negatively
associated with job turnover intention. Strategies such as building a culture of trust
and transparency, offering incentives, providing constructive criticism and
professional development opportunities, may be beneficial. Although the
satisfaction with life scale was not statistically significant in this study, work-life
balance has been identified as a highly critical component to retention rates.
Companies must continue to engage in work-life balance initiatives.
From a theoretical perspective, this landmark study established
that a statistical relationship exists between technology stress and job turnover in
the accounting profession. Future research should seek to determine causality
between technostress and job turnover using structural equation modeling. A
longitudinal study could be performed to measure the perceived variations in
technostress across the accounting profession. A final area of future research that
should be piloted is a cross-comparison of technostress among non-related
professions.
Limitations of the study include the use of a Likert-scale survey as a data
collection tool. Respondents were not provided with the opportunity to expand
upon or explain their answers. The study was limited to the accounting profession.
Results may differ in other occupations or industries. A survey panel was used to
gather data. Accountants not registered as part of the panel may have different
experiences than those surveyed for this study.
CONCLUSION
The accounting profession is poised for monumental change,
driven by advances in technology. Firms must carefully position themselves to
embrace the transformation without further driving up job turnover rates. Through
an awareness of and by implementing approaches to reduce technostress, the
accounting profession can be successful. The workforce will benefit from a more
technostress-conscious workplace.
Journal of Business and Accounting
59
REFERENCES
Agervold, M. (1987). New technology in the office: Attitudes and consequences.
Work & Stress, 1(2), 143-153.
Allen, N. J., & Meyer, J. P. (1996). Affective, continuance, and normative
commitment to the organization: An examination of construct validity.
Journal of Vocational Behavior, 49, 252-276.
American Institute of Stress. (2007). Job stress. Retrieved from
http://www.stress.org/job.htm
Asiedu-Appiah, F., Mehmood, A., & Bamfo, B. A. (2015). Work-life balance
practices, job performance and turnover intentions. International Journal
of Arts & Sciences, 8(4), 379-404.
Babbie, E. R. (2010). The practice of social research (12th ed.). Belmont, CA:
Wadsworth Cengage.
Bao, B., Bao, D., & Vasarhelyi, M. (1986). A stochastic model of professional
accountant turnover. Accounting, Organizations, and Society, 11, 289-
296.
Bedeian, A. G., & Armenakis, A. A. (1981). A path-analytic study of the
consequences of role conflict and role ambiguity. Academy of
Management Journal, 24, 417-424.
Benz, M., & Frey, B. S. (2008). Being independent raises happiness at work.
Swedish Economic Policy Review, 11, 95-134.
Boshoff, C., & Mels, G. (1994). A causal model to evaluate the relationships
among supervision, role stress, organizational commitment and internal
service quality. European Journal of Marketing, 29(2), 23-42.
Brillhart, P. E. (2004): Technostress in the workplace: Managing stress in the
electronic workplace. Journal of American Academy of Business,
Cambridge, 5, 302-307.
Brod, C. (1984). Technostress: The human cost of the computer revolution.
Boston, MA: Addison-Wesley.
Chen, G., Ployhart, R. E., Cooper Thomas, H., Anderson, N., & Bliese, P. D.
(2011). The power of momentum: A new model of dynamic relationships
between job satisfaction and turnover intentions. Academy of Management
Journal, 54, 159-181.
Chong, V. K., & Monroe, G. S. (2015). The impact of the antecedents and
consequences of job burnout on junior accountants’ turnover intentions: A
structural equation modelling approach. Accounting and Finance, 55, 105-
132.
Chua, S. L., Chen, D. T., & Wong, A. F. L. (1999). Computer anxiety and its
correlates: A meta-analysis. Computers in Human Behavior, 15(5), 609-
623.
Dale, K., & Fox, M. L. (2008). Leadership style and organizational commitment:
Mediating effect of role stress. Journal of Managerial Issues, 20(1), 109-
130.
Boyer-Davis
60
Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The Satisfaction
with Life Scale. Journal of Personality Assessment, 49(1), 71-75.
Doll, B. F. (1983). Staff turnover: How to manage it? Journal of Accountancy, 161,
76-82.
Drew, J. (2018). How AI, blockchain, and automation will reinvent accounting.
Journal of Accountancy. Retrieved from
https://www.journalofaccountancy.com/podcast/ai-blockchain-
automation-reinventing-accounting.html
Durndell, A., & Haag, Z. (2002). Computer self-efficacy, computer anxiety,
attitudes towards the Internet and reported experience with the Internet, by
gender, in an East European sample. Computers in Human Behavior, 18,
521-535.
Faul, F., Erdfelder, E., Lang, A. G., & Buchner, A. (2007). G*Power-3: A flexible
statistical power analysis program for the social, behavioral, and
biomedical sciences. Behavior Research Methods, 39(2), 175-191.
Felps, W., Mitchell, T. R., Heckman, D. R., Lee, T. W., Holtom, B. C., & Harman,
W. S. (2009). Turnover contagion: How coworkers’ job embeddedness
and job search behaviors influence quitting. Academy of Management
Journal, 52: 545-561.
Field, A. (2013). Discovering statistics using SPSS (3rd ed.). London, UK: Sage.
Fogarty, T. J., Singh, J., Rhoads, G. K., & Moore, R. K. (2000). Antecedents and
consequences of burnout in accounting: Beyond the role stress model.
Behavioral Research in Accounting, 12, 31-67.
Greenberg, J. (2005). Managing behavior in organizations, (4th ed). Upper Saddle
River, NJ: Pearson.
Greenhaus, J. H., Collins, K. M., & Shaw, J. D. (2003). The relation between work-
life balance and quality of life. Journal of Vocational Behavior, 63, 510-
531.
Gregson, T. (1992). An investigation of the causal ordering of job satisfaction and
organizational commitment in turnover models in accounting. Behavioral
Research in Accounting, 4, 80-95.
Grzywacz, J. G., Butler, A. B., & Almeida, D. (2009). Work, family, and health:
Work-family balance as a protective factor against stresses of daily life.
In The changing realities of work and family: A multidisciplinary
approach (pp. 194-215). Hoboken, NJ: Wiley-Blackwell.
Harrell, A. (1990). A longitudinal analysis of large CPA firm auditors’ personnel
turnover. Advances in Accounting, 8, 233-246.
Howell, D. C. (2007). The analysis of missing data. In W. Outhwaite, & S. Turner
(Eds.), Handbook of social science methodology (pp. 336-346). London,
UK: Sage.
Huselid, M. A. (1995). The impact of human resource management practices on
turnover, productivity, and corporate financial performance. Academy of
Management Journal, 30(4), 721-743.
Journal of Business and Accounting
61
Kahn, R. L., Wolfe, D. M., Quinn, R. P., Snock, J. D., & Rosenthal, R. A. (1964).
Organizational stress: Studies in role conflict and role ambiguity. New
York, NY: Wiley.
Kim, S. S., Im, J., & Hwang, J. (2015). The effects of mentoring on role stress, job
attitude, and turnover intention in the hotel industry. International Journal
of Hospitality Management, 48, 68-82.
Lambert, E. G., Hogan, N. L., Elechi, O. O., Jiang, S., Laux, J. M., Dupuy, P., &
Morris, A. (2009). A further examination of antecedents of correctional
staff life satisfaction. The Social Science Journal, 46, 689-706.
Lee, B. H., & Jamil, M. (2003). An empirical study of organizational commitment:
A multi-level approach. Journal of Behavioral and Applied Management,
4(3), 176-189.
Lee, S. H., & Jeong, D. Y. (2017). Job insecurity and turnover intention:
Organizational commitment as mediator. Social Behavior and
Personality: An International Journal, 45(4), 529-536.
Lee, S. H., Shin, Y., & Baek, S. I. (2017). Task characteristics and work
engagement: Exploring effects of role ambiguity and ICT presenteeism.
Sustainability, 9(10), 1-13.
Locke, E. A. (1976). The nature and causes of job satisfaction. In Handbook of
Industrial and Organizational Psychology. Chicago, IL: Rand McNally.
Linkedin. (2018). Linkedin data analysis reveals the latest turnover trends.
Retrieved from https://business.linkedin.com/talent-solutions/blog/trends-
and-research/2018/the-3-industries-with-the-highest-turnover-rates
Mathieu, J. E., & Zajac, D. M. (1990). A review and meta-analysis of the
antecedents, correlates, and consequences of organizational commitment.
Psychological Bulletin, 108(2), 171-194.
Mitra, A., Jenkins, G. D., & Gupta, N. (1992). A meta-analytic review of the
relationship between absence and turnover. Journal of Applied
Psychology, 77(6), 879-889.
Mitrovska, S., & Eftimov, L. (2016). Calculating the cost for employee turnover
in the IT industry in Macedonia by using a web calculator. Journal of
Human Resource Management, 19(1), 24-33.
Moses, O. S. (2013). Minimizing the effects of technostress in today’s
organization. International Journal of Emerging Technology and
Advanced Engineering, 3(11), 649-658.
Mowday, R. T., Porter, L. W., & Steers, R. M. (1979). The measurement of
organizational commitment. Journal of Vocational Behavior, 14, 224-247.
Nouri, H. (2017). Turnover in public accounting firms: The effect of position,
service line, ethnicity, and gender. Review of Business, 37(2), 14-27.
O’Reilly, C. (1989). Corporations, culture and commitment: Motivation and social
control in organizations. California Management Review, 31(4), 9-25.
Parker, R. J., & Kohlmeyer, J. M. (2005). Organizational justice and turnover in
public accounting firms: A research note. Accounting, Organizations and
Society, 30(4), 357-369.
Boyer-Davis
62
Pasewark, W. R., & Strawser, J. R. (1996). The determinants and outcomes
associated with job insecurity in a professional accounting environment.
Behavioral Research in Accounting, 8, 91-113.
Pew Research Center (2016). Public predictions for the future of workforce
automation. Retrieved from
http://www.pewinternet.org/2016/03/10/public-predictions-for-the-
future-of-workforce-automation/
Pitman, M. K., Gaertner, J. F., & Hemmeter, P. (2011). A study of male and female
employee turnover in public accounting firms. Journal of Applied
Business Research, 5(3), 10-15.
Pucci, E., Cristina, S., Antonaci, F., Costa, A., Imbriani, M., & Taino, G. (2015).
Technostress and primary headache: Psychosocial risk. The Journal of
Headache and Pain, 16(1), 29-31.
Riedl, R., Kindermann, H., Auinger, A., & Javor, A. (2012). Technostress from a
neurobiological perspective: Systems breakdown increases the stress
hormone cortisol in computer users. Business & Information Systems
Engineering, 4(2), 61-69.
Schieman, S., Milkie, M. A., & Glavin, P. (2009). When work interferes with life:
Work-nonwork interference and the influence of work-related demands
and resources. American Sociological Review, 74(6), 966-988.
Senatra, P. T. (1980). Role conflict, role ambiguity, and organizational climate in
a public accounting firm. The Accounting Review, 55, 594-603.
SHRM, (2014). Turnover cost calculator. Retrieved from
https://www.shrm.org/resourcesandtools/tools-and-samples/hr-
forms/pages/1cms_011163.aspx
Siu, O. L. (2003). Job stress and job performance among employees in Hong Kong:
The role of Chinese work values and organizational commitment.
International Journal of Psychology, 38(6), 337-347.
Spector, P. E. (1985). Measurement of human service staff satisfaction:
Development of the Job Satisfaction Survey. American Journal of
Community Psychology, 13(6), 693-713.
Tabachnick, B. G.. & Fidell, L. S. (2013). Using multivariate statistics (6th ed.).
New York, NY: Pearson Education, Inc.
Tarafdar, M., Tu, Q., Ragu-Nathan, B., & Ragu-Nathan, T. (2007). The impact of
technostress on role stress and productivity. Journal of Management
Information Systems, 24(1), 301-328.doi:10.2753/MIS0742-1222240109
Tarafdar, M., Tu., Q., Ragu-Nathan, T. S., & Ragu-Nathan, B. S. (2011). Crossing
to the dark side: Examining creators, outcomes, and inhibitors of
technostress. Communications of the ACM, 54(9), 113-120.
Tett, R. P., & Meyer, J. P. (1993). Job satisfaction, organizational commitment,
turnover intention and turnover: Path analyses based on meta-analytic
findings. Personnel Psychology, 46(2), 259-293.
Toker, B. (2012). Life satisfaction among academicians: An empirical study on the
Universities of Turkey. Social and Behavioral Sciences, 47, 190-195.
Journal of Business and Accounting
63
Tu, Q., Ragu-Nathan, B., & Ragu-Nathan, T. S. (2001). A path analytic study of
the antecedents of organizational commitment of IS managers.
Information Resources Management Journal, 14(3), 27-36.
U.S. Bureau of Labor Statistics. (2018). Job openings and labor turnover,
September 2018 report. Retrieved from https://www.bls.gov/jlt/
Vakola, M., & Nikolaou, I. (2005). Attitudes towards organizational change: What
is the role of employees’ stress and commitment? Employee Relations,
27(2), 160-174.
Viator, R. E. (2001). The association of formal and informal public accounting
mentoring with role stress and related job outcomes. Accounting,
Organizations and Society, 26, 73-93.
Walsh, J. P., Ashford, S. J., & Hill, T. E. (1985). Feedback obstruction: The
influence of the information environment on employee turnover. Human
Relations, 38(1), 23-46.
Wright, K. B., Abendschein, B., Wombacher, K., O’Connor, M., Hoffman, M.,
Dempsey, M., Krull, C., Dewes, A., & Shelton, A. (2014). Work-related
communication technology use outside of regular work hours and work
life conflict: The influence of communication technologies on perceived
work life conflict, burnout, job satisfaction, and turnover intentions.
Management Communication Quarterly, 28(4), 507-530.
Journal of Business and Accounting
Volume 12, No 1; Fall 2019
64
THE COMPREHENSIVE TAXATION SYSTEM
EXISTING DURING THE ROMAN EMPIRE
Darwin L. King
Carl J. Case
Jared L. Roosa
St. Bonaventure University
ABSTRACT
This paper reviews the various forms of Roman taxes that existed during
the period of the Roman Empire. The first section of this paper reviews the
tremendous growth of the Roman Empire over several hundred years. In order to
fund the expansion of the empire and fund the armies, Roman rulers utilized a wide
variety of taxes. The intent of this paper is to review the extremely large amount
of taxes that were extracted from inhabitants of Roman conquered lands. Roman
taxes were in addition to religious taxes paid by Jewish inhabitants of all Roman
controlled lands. This created a system of double taxation which became very
oppressive to those affected.
The first portion of this paper briefly traces the growth of the Roman
Empire over the period from 753BC to 476AD. The major portion of this paper,
however, reviews the Roman taxation system in existence during the years that the
empire existed. Following the brief summary of the growth of the Roman Empire,
but prior to an analysis of the Roman taxation system, is a discussion of the
occupation of the “tax collector” or “publican.” The occupation of tax collector or
“tax farmer,” during this period, was one of very low social status equal to that of
prostitutes and other unsavory vocations. The final portion of this paper reviews
the question of whether or not the inability to assess and collect these various taxes
may have contributed to the fall of the Roman Empire.
Key words: Roman empire, comprehensive taxation system, Romulus and
Remus, Julius Caesar
INTRODUCTION
One of the most impressive civilizations in history is that of the Romans.
Most historians trace the Roman Empire over a period of many centuries.
Depending of the historian, the Roman Empire began with the founding of Rome
by the twin brothers Romulus and Remus (Sidebottom, 2016). As legend has it,
Romulus founded the city on the Palatine Hill. This paper focuses on the taxes that
the Roman government employed to fund all its expenditures including those
Journal of Business and Accounting
65
needed to support the armies. Similar to the taxation system of today, taxes took
many forms including an income tax, a sales tax, various property taxes, and
numerous customs fees and duties.
This paper begins with an overview of the growth and expansion of the
Roman Empire from approximately 753 BC (or BCE) to 476 AD (or CE). The
authors will use the traditional BC (Before Christ) and AD (Anno Domini - Latin
for year of the Lord) in this paper instead of the BCE (Before the Common Era)
and CE (Common Era) designations. Depending on the historian, the period of the
Roman Empire may differ significantly from the dates mentioned above. However,
the authors will adhere to these dates mentioned above. The following paragraphs
discuss the growth of the Roman Empire.
ORIGIN AND GROWTH OF THE ROMAN EMPIRE
During very early times, many Roman citizens believed that Rome was
founded precisely in 753 BC (Sidebottom, 2016). This is the date of the traditional
story when the twins, Romulus and Remus, sons of the god Mars, were placed in
a basket on the river Tiber and left to die. The basket landed on the riverbank at
the future location of the city of Rome. The babies were nursed by a female wolf
and later raised by a local shepherd.
Sidebottom continues the Romulus/Remus myth that disclosed that upon
adulthood Romulus founded the city of Rome on Palatine Hill and killed his
brother Remus. The story is very similar to the Bible’s Old Testament story of Cain
and Able where brother also killed brother. Although Sidebottom admits that there
is no basis in fact for the Romulus and Remus story, he states that archeologists
have found evidence of a city on Palatine Hill as early as 1,000 BC (Sidebottom,
2016). Although an exact date cannot be accurately determined when the city of
Rome first appeared, historian agree that it was at approximately this date.
N.S. Gill provides a brief summary of the periods of history in ancient
Rome (Gill, 2017). He identifies these periods as Regal Rome, Republican Rome,
the Roman Empire, and the Byzantine Empire. The following paragraphs briefly
summarize each of these periods which begin in 753 BC and conclude in 476 AD
(or as late as 1453 AD according to some historians).
Gill states that the Regal Period lasted from 753 BC to 509 BC. This was
a period when kings ruled over Rome with the first being Romulus. Romulus,
according to legend, was the son of the god Mars and a Vestal Virgin named Rhea
Silvia (Gill, 2017). Gill believes that the brothers, Romulus and Remus, were
raised by a prostitute rather than a she-wolf, since the Latin word for brothel is
“lupanar.” The Latin for both prostitute and she-wolf is “lupa.”
Gill also describes a second version of the founding of Rome. The Roman
poet Virgil describes, in the Aeneid, the founding of the town of Lavinium by
King, Case and Roosa
66
Aeneas in approximately 1176 BC. Ascanius, son of Aeneas, built the city of Alba
Longa in 1152 BC. Both these cities were located where the future city of Rome
would be built.
Gill adds that most of what is known about early Rome comes from the
Roman historian, Titus Livius (or Livi) (Livius, 1928). He lived from 59 BC to 17
AD and authored the text “History of Rome From Its Foundation.” Both six and
two volume sets of this work are available online (public domain).
As mentioned above, during the Regal period of 753 BC to 509 BC, Rome
was ruled by seven Kings. Gill adds that legend has it that the Seven Hills of Rome
are associated with these seven original kings. Romulus was king from 753-715
BC. He was followed by Numa Pompilius (715-673 BC), Tullus Hostilius (673-
642 BC), Ancus Martius (642-617 BC), L. Tarquinius Priscus (616-579 BC),
Servius Tullius (578-535 BC), and finally Tarquinius Superbus (Tarquin the
Proud) (534-510 BC). Legend has it that Superbus came into power following the
assassination of Servius Tullius. Superbus and his family were purported to be so
evil and vicious that legend states they were forcibly removed from power by
Brutus and other members of the Senate (Gill, 2017). During this period, the kings
answered to the Senate and did not possess absolute power.
Gill identifies the second period in Roman history as the Roman Republic
or Republican Rome which lasted from 509 BC to 27 BC (Gill, 2017). During this
period, Rome elected the governor of the various regions. To prevent abuse of
power, the Romans allowed the “comitia centuriata” to elect a pair of top officials
who were called “consuls.” The comitia centuriata was a political assembly
comprised of 30 kinship groups into which Roman families were divided. These
kinship groups (called curiae) were comprised of three families that existed when
Romulus was king. These tribes were the Ramnenses, Titienses, and Luceres (Gill,
2017).
The consuls were elected Roman magistrates during the Roman
Republican Period (Gill, 2018). By 181 BC, these men had to be at least 43 years
old and of good character. These individuals possessed a limited term of near
absolute power. They were, however, less powerful than the king because their
term was limited to one year and the power was split between the two consuls.
Consuls were very powerful, since they were responsible for decisions on war,
justice, and finance. Each consul had the power to negate the other. They were,
however, supposed to abide by the senate’s advice.
One of the most important figures during the Roman Republic period was
Julius Caesar who was born in 100BC and lived until he was assassinated on March
15, 44 BC by Marcus Brutus. According to History.com authors, the assassination
conspiracy may have included as many as 60 of his senators who were angry about
the prospect of taking orders from Caesar’s subordinates. These individuals were
Journal of Business and Accounting
67
scheduled to replace Caesar while he was off fighting a war scheduled to begin on
March 18th (History.com authors, 2009).
The third period of the history of Rome, the Roman Empire or Imperial
Rome, according to historiography convention, dates from 27 BC to 284 AD
(Encyclopedia Britannica, 2018). During the period, the Roman Empire spread at
an astounding rate. According to Joshua Mark, the Roman Empire, at its height in
117 AD was the most extensive political and social structure in western civilization
(Mark, 2018). The Roman Empire was so vast in 285 AD that Emperor Diocletian
divided it into a Western and Eastern Empire. The Eastern Empire remained strong
(after 476 AD) operating out of its capital city of Constantinople created by
Emperor Constantine in 330 AD.
This “two headed” empire (Rome and Constantinople) was firmly
established by 395 AD. During this period, Rome was extremely powerful
controlling the Mediterranean, the Balkans, Turkey, and the modern areas of the
Netherlands, southern Germany, France, Switzerland, and England (Gill, 2017).
The Western Empire was dissolved in 476 AD when Rome fell to Germanic King
Odoacer. Some authors consider this date to be the close of the Roman Empire
while others consider the end of the empire to be nearly 1,000 years later.
The fourth period of the history of Rome centered upon the Eastern Empire
referred to as the Byzantine Empire (Mark, 2018). As mentioned above, the capital
city was Constantinople which is modern-day Istanbul. When King Odoacer seized
Rome in 476 AD, he did not destroy the Roman Empire in the east which is known
today as the Byzantine Empire (Gill, 2017). Residents of this area spoke either
Greek or Latin but were citizens of the Roman Empire. However, the culture in
this area was more Greek than Roman.
Although the Western Empire collapsed in 476 AD, the Eastern Empire
continued to exist for approximately 1,000 years. It was not until the Ottoman
Turks conquered the area in 1453 AD (Gill, 2017) Using this date, most historians
identify the timeline for the Roman Empire to begin in 753 BC and end in 1453
AD. This 2,200 year period marked the beginning and end of one of the most
recognized and respected civilizations in our history.
The remainder of this paper will review several types of taxes assessed
upon residents of the Roman Empire. A comprehensive tax system was essential
in order to finance the cost of the Roman armies and provide services for all
inhabitants of the vast empire. Prior to the review of the Roman tax system, a
description of the duties and responsibilities of a publican or tax farmer (collector)
is appropriate.
King, Case and Roosa
68
THE PROFESSION OF PUBLICAN OR TAX COLLECTOR (FARMER)
The word “publican” is an English translation of the Greek word “telones”
which means “tax-farmer” (Badian, 1983). Therefore, a publican had the task of
collecting taxes for the Roman government. During the period of the expansion of
the Roman Empire, these individuals would bid to acquire a “tax franchise” for a
particular area that enabled them to collect a variety of taxes from occupants of the
area. These tax farmers were responsible for converting taxes collected in-kind
(crops/animals) into hard currency. Therefore, publicans collected enough money
from taxpayers to cover the taxes due, exchange fees, and provide themselves with
a profit. (Bartlett, 1994). In addition to an already extravagant salary, the publicans
would collect additional taxes above those required by Rome which would further
increase their total income.
During the first century AD, Jerusalem was under Roman control being
conquered by Roman general Pompey during his eastern campaign when he
established the Roman province of Syria in 64 BC and conquered the city of
Jerusalem in 63 BC (Barraclough, 1981). In this region, publicans were typically
Jews who worked for the despised Roman government collecting various taxes
from Jewish citizens. Publicans were despised in most cultures, since they were
automatically assumed to be dishonest, unethical, and without character.
Basically, the invading Roman government employed citizens of
conquered lands to perform the undesirable task of collecting taxes. The Roman
government accomplished this by promising significant bonuses. This allowed the
publicans to extort as much money as possible from the local citizens thereby
increasing the publican’s total income. It is easy to imagine the significant
corruption that existed in this type of tax collection system. It is also
understandable that the publicans were despised and deemed to be traitors to their
own nation. Given this situation, publicans had few friends and were forced to
associate with other publicans or the criminal element of the population. The Bible
states, in many of its books, that association with a publican automatically cast
suspicion on that person’s reputation.
The Bible reports that Jesus was often found in the company of publicans.
For example, one of the disciples, Matthew (also called Levi) was a tax collector.
“And as Jesus passed forth from thence, he saw a man, named Matthew, sitting at
the receipt of custom: and he saith unto him, Follow me. And he arose, and
followed him” (Matthew 9:9). This appalled the religious Jewish leaders as verse
11 continues with, “And when the Pharisees saw it, they said unto his disciples,
Why eateth your Master with publicans and sinners?” The Bible contains many
verses where tax collectors are on the same level as prostitutes and other
undesirables such as Matthew 21:32 (publicans and prostitutes (harlots)), Luke
5:30 (publicans and sinners), and Mark 2:16 (publicans and sinners).
Journal of Business and Accounting
69
John MacArthur also comments on the social status of the publican. He
discusses the disciple Matthew who possessed the occupation least likely to
warrant discipleship. MacArthur states that tax collectors were the most despised
people in Israel (MacArthur, 2002). He states the citizens of that period deemed
them to be lower and more worthy of scorn that the occupying Roman Army.
MacArthur reports that publicans extorted money from the people often using hired
thugs. In his words, publicans were considered “despicable, vile, unprincipled
scoundrels.” MacArthur states that publicans had an unspoken agreement with the
Roman Emperor that besides collecting legitimate taxes owed to Rome that they
could also assess additional fees and taxes. These additional monies could be
retained by the publican in full or shared with Rome on a percentage basis.
MacArthur reports that there were two types of tax collectors known as
the “Gabbai” and the “Mokhes” (MacArthur, 2002). The first group (Gabbai) were
general tax collectors who collected property, income, and poll taxes. The second
type (Mokhes) collected duties on imports and exports, goods for domestic trade,
and virtually everything that was moved by road. In addition, they established tolls
on roads and bridges, taxes on beasts of burden, and taxes on transport wagons
based on the number of axles. Finally, MacArthur added that publicans were
allowed to assess taxes on packages/parcels, letters, and any other type of personal
property.
There were two types of Mokhes referred to as the Great Mokhes and the
Little Mokhes (McArthur, 2002). A Great Mokhes operated behind the scenes and
hired others to actually collect the taxes in his place. A Little Mokhes operated a
tax office where he would deal with people on a face to face basis. MacArthur
believes that the disciple Matthew was a Little Mokhes because he was dealing
with people face-to-face at a tax office when Jesus found him (Matthew 9:9).
MacArthur stresses how much publicans were hated by the local people. He states
that no self-respecting Jew would become a tax collector. This action cut him off
from his people and his God, since he was banned from the synagogue and
forbidden to sacrifice and worship in the temple. As mentioned earlier, the
publicans had few friends other than fellow publicans and the socially undesirable.
George Shillington discusses the “Tale of Two Taxations” (Shillington,
1997). This relates to taxation at the time of Christ during the first century AD.
Inhabitants of Jerusalem during this period were required to pay both Roman and
Jewish taxes which created a dual tax system. The publicans were especially hated
because they extracted money from Jewish residents that was budgeted for one of
their three required religious taxes. These included a wave or first fruits offering
of typically 3%, a 10% annual tithe (for support of the priests and Levites), and a
second 10% tithe for a total of approximately 23%. The second portion of the dual
tax system included the Roman taxes collected by the publicans. Shillington
references Titus Flavius Josephus who wrote that the primary taxes extracted by
the Romans were the poll (head tax) and land taxes. In addition to these two taxes,
King, Case and Roosa
70
the Roman government also collected various land transfer taxes, export and
import duties, and a real estate tax on all the houses in Jerusalem (Josephus,
78AD). It is easy to understand the hatred towards the Publicans who extracted
numerous taxes to finance a foreign government. This left residents of conquered
lands with less funds to pay necessary living expenses and pay required Jewish
religious taxes identified in both the books of the old and new testaments of the
Bible.
The next portion of this paper identifies the significant number of taxes
that were utilized by the Roman government. These included real estate taxes,
excise and custom taxes, marriage taxes on unmarried men and women who could
not bear children, inheritance taxes, poll taxes, and many more. The final segment
of this paper reviews the question of whether massive tax evasion was a major
reason for the fall of the Roman Empire.
THE EXTENSIVE ROMAN TAXATION SYSTEM
The Roman Empire extracted a number of direct and indirect taxes or
tributes. The two basic types of direct taxes were tributum soli and tributum capitis
(Encyclopedia Britannica, 2018). According to Brunt, tributum soli was a tax on
the land (Brunt, 1981). It was not a tax on harvests or crops. Nor was it a tax on
“movables” such as slaves, animals, and equipment for cultivating and processing
crops. Since land was the chief source of wealth during the period of the Roman
Empire, the land tax produced the majority of total tax revenues.
Caesar Augustus (or Octavian until 27BC), the first Roman emperor,
following the republic, from 30BC to 14AD, established a number of fiscal and
monetary reforms. In particular, he introduced three new taxes including a land tax
(tributum soli), a general sales tax, and a flat-rate poll tax (tributum capitis)
(Davies, 2002). The tributum soli or land tax was established at one percent of the
assessed value of the land. The general sales tax of 1% and flat-rate poll tax on
adults from the age of 12 or 14 to age 65 will be discussed later in this paper.
Brunt states that in the Roman census process, the imperial government
had a model for registration and taxation of all forms of property including both
real and tangible personal property. This included land, buildings, cash, debts due
from others (accounts receivable), clothing, jewels, and slaves (Brunt, 1981). The
worth of all these forms of capital was valued apparently using a set of formulas
established by the censors who were government officials who conducted the
census. This valuation system required the possessors of the land to specify the
acreage that could be cultivated and farmed, the number of grade vines, and the
acreage containing olive trees (Brunt, 1981).
Since the land tax was such a vital part of the Roman taxation system, there
were significant penalties for false declarations concerning the acreage and value
of the property. Slaves implicated in the frauds would often suffer death. The
Journal of Business and Accounting
71
testimony of slaves was often sufficient to convict the land owner. Septimius
Severus, Roman emperor from April 193AD to February 211AD ruled that slaves
could be examined against their owners for only the crimes of census fraud and
adultery (Brunt, 1981).
A second type of Roman tax was called “Collatio lustralis” or a tax on
“traders in the widest sense.” It was a tax on anyone who produces a product or
provides a service, with the exception of farmers, physicians, and teachers (Oxford
Classical Dictionary, 1970). This tax was instituted by Constantine, Roman
Emperor who ruled from 306 to 337AD. According to the Byzantine writer
Zosimus, Constantine established this tax as early as 325AD (Goffart, 1971).
The tax was originally collected annually in either gold or silver but in the
late 4th century it was only payable in gold. In the early 5th century, the tax was
required to be paid once every four years. Zosimus reported the extreme hardship
caused by this tax, since it was collected in a lump sum every four years (Goffart,
1971). It was devastating to parents who were forced into selling their children into
slavery or prostitution in order to pay this tax.
The tax was abolished by Anastasius I in 498AD in the Eastern Roman
Empire as part of his monetary and fiscal reforms (Nicks, 1998). Anastasius was
known for his administrative efficiency. For example, he changed the payment
method for governmental transactions from goods to hard currency (Treadgold,
2001). Upon leaving the imperial government, he was recognized for establishing
a significant budget surplus due to internal controls introduced to minimize
governmental corruption. The state treasury, at the time of his death, was three
hundred and twenty thousand pounds (Nicks, 1998). In addition, Anastasius
reformed the tax code of the period and introduced an improved form of currency
(Treadgold, 2001).
A third type of Roman tax was called Portoria which was also referred to
as a harbor or customs tax. This was a 2½% ad valorem (based upon value) tax
collected in virtually every port in the Mediterranean on nearly every type of
product (Finley, 1973). The only items exempted from the tax were shipments of
imperial annona (corn supplies) for the armies and the citizens of Rome.
The system supplying Rome, the armies, and other valuable cities with
grains (mainly corn) and other foodstuffs came to be known as the annona
(Erdkamp, 2016). The grain distribution system was first introduced in 123BC by
C. Sempronius Gracchus (a Roman populist and reformist politician). Given that
the harbor tax or customs duty was assessed on virtually all goods moving in and
out of Mediterranean ports, this tax produced significant revenue for the Roman
government. This system ended in 429AD when Africa was conquered by the
Vandals, since most grains at this time were shipped to Rome from African ports.
King, Case and Roosa
72
The Roman tax system also included two taxes assessed on orphans,
widows, and single women that was used for the purchase and upkeep of the
Roman army’s horses. These taxes were termed “aes equestre” and “aes
hordearium” (Hill, 1943). The aes equestre was the monetary sum of 10,000 asses.
The as (plural asses) was originally introduced in approximately 280BC as a large
cast bronze coin (later copper) that was minted during the periods of the Roman
Republic and Roman Empire.
The aes equestre was given to a cavalry soldier often called the “equus
publicus” (equestrian order) for the purchase of a horse. The equus publicus was a
wealthy secondary property-based class of ancient Rome who ranked below the
senatorial class. In more modern times, this class would have been known as
“knights.” Therefore, the cavalry was composed of men from fairly affluent
heritage.
The aes equestre was collected from single women (maidens and widows)
and orphans provided that they owned a certain amount of property (Smith, 1875).
This was based on the argument that women and children should contribute to the
military who fight on their behalf. The Roman government deemed this to be a fair
tax given the fact that women and children were not included in the census and
therefore were exempt from other taxes such as the poll tax which was assessed
primarily on adult men (Niebuhr, 1835). Revenue from this tax was deposited and
accumulated in the public treasury. Later, these revenues were paid from the public
treasury to a cavalry soldier in lots of 10,000 asses in order to purchase a horse
(Smith, 1875).
The aes hordearium amounted to the sum of 2,000 asses annually for the
care and feeding of a horse (Smith, 1875). Like the aes equestre, the tax was
assessed against single women and orphans who owned a certain amount of
property. In other words, it was not collected from women and orphans who had
little or no assets. The cavalry soldier received this annual 2,000 asses stipend from
the public treasury at the rate of 200 asses per month for 10 months (Niebuhr,
1835).
Niebuhr states that in these early times ten months were considered to be
one year and 10 months times the 200 asses per month equals the aes hordearium
total of 2,000 asses for a year. He continues that the cavalry rate of 200 asses per
month was twice the 100 asses received by a foot-soldier or infantryman (Niebuhr,
1835). Niebuhr adds that generals in the Roman Army received 300 asses per
month. The general’s triple pay (triple the infantryman’s salary of 100 asses) was
introduced in 354AD by the military tribune Cn. Cornelius Cossus (Niebuhr,
1835).
Another Roman tax was called the “aes uxorium” which was a tax imposed
upon women who could not bear children, with the exception of Vestal Virgins,
Journal of Business and Accounting
73
and unmarried men who had reached adulthood without marrying (Peck, 1898). It
was first introduced by the censors M. Furius Camillus and M. Postumius in
403BC (Smith, 1875). This tax did not apply to unmarried men age 60 and older
nor to unmarried women age 50 or more. It is uncertain how long this tax remained
in existence.
Prior to the reign of Rome’s first emperor, Augustus Caesar, from 27BC
to 14AD, no inheritance taxes were recorded for the Roman Republic. Augustus
introduced the “vicesima hereditatium” or inheritance tax of approximately 5%
(Gardner, 2001). This tax applied to inheritances that were received by way of a
will and any close relatives were exempt from paying it. Close relatives included
the deceased individual’s grandparents, parents, children, grandchildren, and
siblings. Roman custom from the late Roman Republic and afterwards was that
husbands and wives kept their own property separate from that of the spouse. This
resulted in Roman women remaining a part of her birth family. She was not
considered to be under the legal control of her husband. Therefore, in many cases
it was questionable if the wife was exempt from the tax. Many authors feel that
Roman social values regarding marital devotion probably exempted the wife from
tax in most cases (Gardner, 2001).
Another tax that was introduced by Augustus Caesar was a 1% general
sales tax (Davies, 2002). The tax was designed to fund the military and was
instituted in approximately 6AD. It was termed the “centesima rerum venalium”
which roughly translates to “a hundredth of the value of everything sold.” This tax
was later reduced to one half percent by Tiberius and finally eliminated by
Caligula. Tiberius was emperor from 14AD to 37AD following Augustus. Caligula
followed Tiberius as emperor from 37AD to 41AD. This unpopular sales tax
existed for less than 40 years.
A Roman tax that was specifically imposed upon Jewish people living in
the Roman Empire was called the “fiscus Judaicus” (Gottheil & Krauss, 1906).
According to Josephus, the tax was imposed by Roman Emperor Vespasian as a
result of the First Roman-Jewish War (first Jewish revolt) of 66-73AD. The tax
was imposed following the destruction of the second Temple in 70AD. This two
denarii (or one-half shekel) tax was in place of the Temple tax (Exodus 30:vs.11-
13) paid in prior years by the Jewish people for the upkeep of the Temple. This tax
was a major humiliation for the Jews who had just experienced the destruction of
their Temple for the second time. Unlike the Temple tax of previous years that was
paid only by adult men between the ages of 20 and 50, the fiscus Judaicus was
imposed on all Jewish people including women, children and the elderly (Schäfer,
1998).
The extensive Roman tax system also included a poll or head tax (of two
denarii) called the “tributum capitis” (Digest 50, tit.15, 1932). The Roman
government imposed this direct tax on all the people living in the controlled
King, Case and Roosa
74
provinces. This tax was assessed primarily on the Roman subjects living in the
provinces and not on Roman citizens. For this reason, it was a much hated tax
referred to by Tertullian (an early Christian author) as a "badge of slavery." (Audi,
(1999). According to Tertullian, this tax provoked numerous civil revolts. The
most famous of these being the Zealot revolt in Judea in 66AD following the
destruction of the temple in 70AD. These taxes were typically collected by the
private tax farmers or publicans discussed earlier. According to Tertullian, these
taxes were assessed annually on males over age 14 and females over age 12 until
they reached the age of 65.
The final two Roman taxes included in this paper are typically called slave
taxes (Bradley, 2016). The “vicesima libertatis” was a tax on slave owners who
freed slaves. The tax was 5% of the value of the slave (Dilke, 2016). The tax was
either paid by the owner, if he freed the slave, or by the slave, if he redeemed
himself with his own money. The second slave related tax was called the “quinta
et vicesima venalium mancipiorum.” This was a 4% tax on the owners who sold
slaves (Dilke, 2016). Information is very scarce concerning these taxes such as the
years they existed and the methods used to value the slaves. No doubt age and sex
were important assessment factors but details on this valuation process have been
lost over the centuries.
It is obvious that the Romans were forced to utilize a wide range of taxes
in order to finance the expansion of the empire and purchase food, supplies, and
armaments for the armies. In addition, funds were needed to purchase food for the
large population (approximately one million) living in the city of Rome. Davies
estimates that Rome needed to import at least 150,000 tons of grain each year to
feed the inhabitants (Davies, 2002). Much of this grain was needed to feed the poor
of the city. Davies states that as many as 200,000 of the poor in Rome received
free distributions of wheat from the Roman government. It is not surprising that
the cost of funding the armies and expanding the empire required the emperors to
consider all possible forms of additional tax revenues.
A number of Roman emperors revised and modified the monetary and tax
system for Rome. One of the most successful in these endeavors was Augustus
Caesar emperor from 30BC to 14AD. He instituted a set of three principal taxes to
generate the much needed revenues for the period (Davies, 2002). As mentioned
earlier, these included a 1% sales tax, a 1% tax on land (tributum soli), and the poll
or head tax of two denarii on men aged 14 to 65 and women aged 12 to 65. Many
of details of the taxes discussed above have be lost over the centuries such as the
exact year introduced and removed. Finally, an interesting fact is that the majority
of the Roman taxes (real estate, sales, etc.) discussed in this paper are still common
today in many countries around the world.
Journal of Business and Accounting
75
FINAL THOUGHTS – TAXES AND THE DECLINE OF THE ROMAN
EMPIRE
No doubt the decline of the Roman Empire was the result of numerous
factors. However, two of the major causes for the decline could have logically been
the widespread tendency of the people to evade as many of the multitude of taxes
as possible and the inefficient and ineffective tax assessment and tax collection
processes of the Roman government.
Pulliam discusses a number of reasons why significantly more taxes were
collect than were actually deposited with the Roman government (Pulliam, 1924).
In addition to taxpayers evading these taxes, there was also a huge amount of fraud
and embezzlement on the part of tax assessment and tax collection personnel.
Pulliam argues that that “while the revenue secured was usually adequate for all
expenses of the government, the system of taxation had a number of serious defects
(Pulliam, 1924).
The primary defect, according to Pulliam, was the method of collecting
the taxes. As discussed earlier in this paper, the publicans (tax farmers) were
considered cheats and fraudsters. Pulliam states that publicans collected exorbitant
amounts of tax well above the required amount. He says that the publican became
the “most odious figure in provincial life and perhaps the most galling feature of
Roman domination was the tax that Rome collected” (Pulliam, p.548).
The second basic problem of the Roman tax system, according to Pulliam
was the range of duty rates charged on goods moved from one province to another.
The trade districts were often arbitrarily determined and did not correspond to
provincial boundaries (Pulliam, p.548). Pulliam states that the duties on goods
going from one province to another were much higher than goods going from a
province to Rome. In addition, harbor duties and road and bridge tolls made the
cost of transporting products from province to province cost prohibitive. Pulliam
argues that free movement of trade throughout the empire would have increased
the overall wealth of the empire significantly.
A final problem of the Roman Empire taxation system, according to
Pulliam, was the lack of overall general administration (Pulliam, p.549). The
Roman government never attempted to appoint an experienced, effective, and
efficient manager to the finance and taxation department of the government. The
head of finance and expenditures, called a “quaestor,” served for only one year
thereby never becoming familiar with the details of the financial management area
of the government. Pulliam states that the budgeting process was nonexistent and
that there was no attempt to balance taxation revenues and government
expenditures.
King, Case and Roosa
76
SUMMARY AND CONCLUSIONS
This paper reviewed many of the taxes collected by the Roman
government during the period of the Roman Empire. It appears that most every
type of tax in existence today is rooted in the history of the Roman Empire which
covered many centuries from 753BC to 476AD. The tax system of this period was
a major hardship on the people living within the empire. This resulted, in large
part, from the fraudulent and dishonest actions of the publicans or tax collectors of
the period. It is easy to understand the importance of internal controls in today’s
business environment. Unfortunately, such controls did not exist during the period
of the Roman Empire.
REFERENCES
Audi, R. (1999). The Cambridge Dictionary of Philosophy. Cambridge
University Press, 908.
Badian, E. (1983). Publicans and Sinners. Cornell University Press
Barraclough, G., ed. (1981). The Times Atlas of World History.
Bartlett, B. (1994). How Excessive Government Killed Ancient Rome, Cato
Journal, 14. 2 (Fall).
Bradley, K. (2016). The vicesima libertatis: Its History and Significance. Klio,
66.66, 75-182. Retrieved, from doi:10.1524/klio.1984.66.66.175.
Brunt, P. (1981). The Revenues of Rome. The Journal of Roman Studies, 71,
161-172. doi:10.2307/299505
Davies, G. (2002). A History of Money, University of Wales Press,
Cardiff, 96. Digest 50, tit.15, 1932.
Dilke, O. A. W. (1987). Mathematics and Measurement, Berkeley:
University of California Press, 51.
ENCYCLOPEDIA BRITANNICA. (2018). ROMAN EMPIRE ANCIENT
STATE [27 bc-476 ad], retrieved from
https://www.britannica.com/place/roman-empire
Erdkamp, P. (2016). Annona (grain), Oxford Classical Dictionary. Retrieved
from http://oxfordre.com/classics/view/10.1093/acrefore/
Finley, M. I. (1973). The Ancient Economy, University of California Press,
Berkeley and Los Angeles, CA.,159.
Journal of Business and Accounting
77
Gardner, J. F. (2001). Nearest and dearest: liability to inheritance tax in Roman
Families, Class and Kin in the Roman World. Routledge, 205-220.
Gill, N.S. (2107). Periods of History in Ancient Rome, Thought Co.
Retrieved from https://www.thoughtco./periods-of-ancient-rome.
Gill, N.S. (2107). Romulus-Roman Mythology About the Founding and First
King of Rome, Thought Co. Retrieved from
https://www.thoughtco.com/romulus-roman-mythology.
Gill, N.S. (2107). Who Were the Early Kings of Rome? Thought Co.
Retrieved from https://www.thoughtco.com/the-early-kings-of-rome
Gill, N.S. (2108). Who Were the Roman Consuls and How Did They Rule
Rome? Thought Co. Retrieved from https://www.thoughtco.com/who-
were-the-roman-consuls
Goffart, W. (1971). Zosimus, The First Historian of Rome's Fall. The
American Historical Review, 76.2, 412-441. doi:10.2307/1858706
Gottheil, R. & Krauss, S., (1906). Jewish Encyclopedia of 1906. Retrieved from
http://jewishencyclopedia.com/articles/6157-fiscus-judaicus on 1/14/19.
Hill, H. (1943). Aes Equestre, Aes Hordearium, and Triplex Stipendium.
Classical Philology, 38.2, 132-134. Retrieved from
http://www.jstor.org/stable/264298
History.com (2009). The ides of March: Julius Caesar is murdered, A&E
Television Networks, November 13.
Holy Bible, King James Online Version. Retrieved from
https://www.kingjamesbibleonline.org
Josephus, T.F. (78AD). The Wars of the Jews, xix.vi.3.
Livius, T. (1928). History of Rome, translated by George Baker, British
Library (Historical Print Editions), originally published in six volumes in
1828 and later published in a two volume set in 1844.
MacArthur, J. (2002). Twelve Ordinary Men, Thomas Nelson Publishing,
151-157.
Mark, J.J. (2018). Roman Empire, Ancient History Encyclopedia.
Retrieved from https://www.ancient.eu/Roman_Empire/
King, Case and Roosa
78
Niebuhr, G.B. (1835). The History of Rome, translated by J.C. Hare and C.
Thirlwall, Volume II, Thomas Wardle, Philadelphia, PA.
Nicks, F.K. (1998). The Reign of Anastasius I, 491-518, St. Hilda’s College,
Oxford.
Oxford Classical Dictionary, 2nd ed. (1970).
Peck, H.T. (1898). Harpers Dictionary of Classical Antiquities, New York.
Harper and Brothers.
Pulliam, R., (1924). Taxation in the Roman State, The Classical Journal, 19.9,
545-553.
Schäfer (1998). Digest 50, 113–114.
Shillington, V.G. (1997). Jesus and His Parables, T&T Clark LTD, Edinburgh,
Scotland, 23-33.
Sidebottom, H. (2016). 10 Key Roman Date You Need to Know, History
Extra, November.
Smith, W. (1875). Aes Equestre, Aes Hordearium, and Aes Militare, A
Dictionary of Greek and Roman Antiquities, John Murray, London, 26.
Treadgold, W. (2001). A Concise History of Byzantium, Houndmills,
Hampshire: Palgrave, 57.
Journal of Business and Accounting
Volume 12, No 1; Fall 2019
79
KEEPING THE OIL AND GAS PIPELINES OPERATING
AS ASSETS: SAFETY ISSUES CAN MAKE THEM
LIABILITIES
Carol Sullivan
University of Texas Permian Basin
ABSTRACT The purpose this research is to examine the pros and cons of oil and gas
pipelines. While most people think of these pipelines as “Assets”, they can become
“Liabilities”. Risk assessment and legal liability issues will be examined. The key
to upholding safety standards is vigilant behavior on the part of the corporate
leaders, operating managers, employees, and even the general public. Observation
and “see something, say something” measures can prevent huge losses – human
life, environmental damage, and monetary payouts. The research is important for
USA energy security concerns, USA jobs, environmentalists, and taxpayers.
Key Words: Risk Assessment, Oil and Gas, Pipelines, Safety
INTRODUCTION While the typical price of a barrel of oil ranged from about $50-70 in 2018,
the estimated cost of a barrel of spilled oil is about $19,800. The oil spill costs at
Marshall, Michigan were as high as $60,000 per barrel in 2010! The purpose this
research is to examine the pros and cons of oil and gas pipelines. While most
people think of these pipelines as “Assets”, they can become “Liabilities” with
mishaps and negligence. The purpose of this research is to examine some of the
incidents in West Texas, the United States, and other parts of the world in order to
educate people about the types of risk taken when dealing with oil and gas
pipelines. While pipelines continue to be the safest mode of product transportation,
it is important to remain vigilant about the risks and ways in which these assets can
turn into liabilities.
LITERATURE REVIEW
According to Global Data (2018), the global oil and gas pipeline industry
is expected to witness the start of the operation of 504 pipelines during the 2015–
2018 period. Of these, 315 will be natural gas pipelines, 105 will be crude oil
pipelines. Makholm (2016) describes that the regulatory treatment of oil
pipelines in the United States has progressed, as the Federal Energy Regulatory
Commission (FERC) attempts to reconcile what it knows about regulating gas
pipelines as a highly competitive transport business with its oil pipeline duties.
Mogel and Gregg (2004) asserts that natural gas now is more expensive than fuel
oil in many parts of the US. This has caused a significant problem for interstate
natural gas pipelines which find that their markets are eroding at the same time
when they must take or pay for large volumes of natural gas because of
contractual obligations incurred in the middle and late 1970s. As a partial
solution, some have argued that the existing approximately 269,000 mile natural
Sullivan
80
gas pipeline transportation system is a barrier to the sale, transportation, and use
of natural gas unless that system is converted to one of common, or at least,
contract carriage. Klass,& Meinhardt (2015) explores the history and geography
of oil and natural gas to help explain why US regulation of the infrastructure for
transporting these two similar types of energy resources to markets developed so
differently. Notably, while interstate natural gas pipelines are reviewed and
permitted at the federal level by the Federal Energy Regulatory Commission
(FERC), interstate oil pipelines are reviewed and permitted almost exclusively at
the state level. This research considers whether changes in the regulatory
agencies for oil or natural gas transportation are now needed to promote more
effective cross-country transportation of new sources of shale oil and gas made
available by hydraulic fracturing technologies.
Dey, et al (2004) describes how offshore oil and gas pipelines are
vulnerable to environment as any leak and burst in pipelines cause oil/gas spill
resulting in huge negative impacts on lives. Breakdown maintenance of these
pipelines is also cost-intensive and time-consuming resulting in huge tangible and
intangible loss to the pipeline operators. This study develops a risk-based
maintenance model whereby risk-based inspection and maintenance methodology
is particularly important for oil pipelines system, as any failure in the system will
not only affect productivity negatively but also has tremendous negative
environmental impact. The research proposes a model that helps the pipelines
operators to analyze the health of pipelines dynamically, to select specific
inspection and maintenance method for specific section in line with its probability
and severity of failure. Glukhova & Zubarev (2017) focuses on an economically
expedient resolution of the problem of industrial and environmental safety of
operation of the existing gas pipelines. It highlights the importance of safety and
reliability of pipeline transportation for the economic stability of the industry and
energy security of the country and demonstrates economic impacts of accidents on
the main gas and oil pipelines. Rigney (2015) Technological advancement in
drilling techniques, primarily hydraulic fracturing, has provided access to
previously unreachable natural gas reserves. Much of this increase in natural gas
production is derived from the Marcellus Shale, a shale formation that spans Ohio,
Pennsylvania, West Virginia, and New York. This surge in natural gas production
has prompted natural gas pipeline companies to upgrade their pipeline networks.
Pipeline companies must apply for certificates of public convenience and necessity
from the Federal Energy Regulatory Commission (FERC) and, if approved,
perform an environmental evaluation, as required by the National Environmental
Policy Act (NEPA). In examining the environmental impacts of the pipeline
project, pipeline companies must be careful not to impermissibly segment the
project into component parts, thereby failing to consider a proposed project's full
range of environmental impacts. This is referred to as the rule against
segmentation, developed by courts to ensure that companies consider the full range
of environmental consequences of proposed projects. A worldwide approach to the
issues can be found in Kolb (2012) as it examines the ongoing natural gas
Journal of Business and Accounting
81
revolution and assesses its impact on the energy industry and societies of Central
Asia. The natural gas revolution consists of three related technological
developments - hydraulic fracturing, horizontal drilling, and the increasing build-
out of the world liquid natural gas (LNG) infrastructure. The article focuses on
Turkmenistan and its rich reserves of natural gas and explores the conditions under
which Turkmenistan currently reaches international markets through pipelines to
China, Iran, and Russia. It also assesses Turkmenistan's future prospects for
reaching additional world markets and for sustaining the markets it presently
serves. Finally, the article analyzes the difficulties that Turkmenistan's gas industry
are likely to face and the implications these continuing energy industry tribulations
will have for social development in Central Asia.
Leitzinger & Collette (2002) identifies lessons learned from the
restructuring of the natural gas industry. The replacement of regulated wholesale
merchants by new entrants did not occur because of competitive efficiencies.
Competition in wholesale gas marketing has produced consumer benefits in the
form of innovation, new product development, broader choices and possibly
reduced costs. Mbara & Van (2011) describes the rapid growth in the use of
pipelines to transport energy products. Due to the strategic nature of energy
products that are transported by pipelines, the importance of risk awareness,
assessment and management cannot be over-emphasized. With the risk of pipeline
disruptions increasing globally, energy pipeline organizations are compelled to
incorporate measures that should help to identify and address areas that can lead
to energy pipeline disruptions. Given the strategic importance of energy pipelines,
the main purpose of this research is to ascertain awareness of the risks associated
with the energy pipeline’s physical environment, not only from the energy pipeline
operators, but also from communities who are exposed to such risks. The findings
show that the corporate energy sector in South Africa is aware of risks associated
with energy pipeline supply chains while the general public’s awareness is very
low.
Jamal (2015) asserts that ensuring the availability of basic
resources and energy services, in adequate quantities and at reasonable prices, is a
vital aspect of national energy policy everywhere. Hence, transnational pipeline
projects for the transportation of oil and gas products play a significant role in the
global quest for energy security. The construction and operation of cross-border
pipeline systems from producing countries to consumer markets entail huge
financial commitments for investors. Likewise, such long-term infrastructure
projects straddling several jurisdictions have implications for the rights of
livelihood-affected people and may adversely impact the environmental
sustainability along the project area. The research pays close attention to the
Energy Charter Treaty, the only multilateral legal instrument dealing exclusively
with the energy sector.
Hellmann (2015) describes how the threat of large-scale cyber
attacks on the nation's oil and gas pipeline systems is increasing. With increased
Sullivan
82
information sharing between the private sector and the government, and specific,
numeric objectives to work toward in developing cybersecurity programs for
pipeline systems, the voluntary measures currently in place might prove effective
in protecting systems nationwide. These voluntary measures could be strengthened
through legislation streamlining the information sharing process and providing
liability and privacy protection for oil and gas pipeline owners, which would
further incentivize industry participation.
In summary, it is important to remember the difference in price
between a product that is transported properly and remains an asset as compared
the product becoming a “Liability”. Just because the federal government is pro-
pipeline, that will not eliminate the risk assessment stage and the legal liability that
can result from mishaps.
TEXAS PIPELINES AND MISHAPS
Activity in the Permian Basin area of Texas has reached levels
higher than any recorded in history and this region is one of the main reason why
the United States is a world oil producer now. There is motivation for the oil
pipeline rivals to consolidate some of the West Texas projects because the lack of
pipeline transports is the major constraint in the area. While consolidation can take
advantage of certain companies’ expertise, there is also risk involved if the
companies do not work together properly in their new roles. There has been an
increase in natural gas flaring simply because the area has insufficient pipeline
infrastructure and flaring causes light pollution for the area. This creates concerns
for the McDonald Observatory enthusiasts in the area.
A new natural gas pipeline is being proposed from southeast New
Mexico to Corpus Christi. The 650-mile pipeline will transport 220,000 barrels per
day of natural gas liquids when completed. Natural gas liquids include propane,
butane, and ethane, which is used in plastics production and as a petrochemical
feedstock. The pipeline will extend from the Permian basin in southeast New
Mexico and West Texas to Corpus Christi, where a fractionation complex will be
built to separate the various liquids.
In addition, eight other companies are exploring the Delaware
Basin. There is a new finding of oil in the area of Balmorhea Lake in southern
Reeves County, about 100 miles from the heart of the Permian Basin oil patch too.
Apache, the oil company involved with it, estimates that the area may contain 3
billion barrels of oil. The company named the new field Alpine High. Further
south, in Big Bend country, Energy Transfer Partners has built two natural gas
pipelines, the Comanche Trail and Trans-Pecos pipelines. "The pipelines are really
starting to lay the groundwork of what made it possible for companies to frack in
these parts of the Permian Basin. The new pipelines come at the best possible time
for Apache, which will need a way to move the oil it produces.
Increased energy production will have an environmental impact
on the region. Increases in fracking in the Permian Basin overall have already
stressed the environment. The main question is whether the country's natural
Journal of Business and Accounting
83
resources and wild spaces will be used for development. Water is a key concern
in the region and the unique spring-fed pool at Balmorhea State Park is of major
concern. It is the west Texas desert oasis that drives tourism in the area, and the
economy.
With all of the production and oil activities, the assets can quickly
be transformed into liabilities when construction is not properly conducted and
when operations go awry. Here are a sample of very recent Texas mishaps that
costs both companies and the state money:
• On January 30, 2017: A Texas Department of Transportation crew dug
into the 30-inch-diameter Seaway Pipeline, near Blue Ridge, Texas,
spraying crude oil across road. About 210,000 gallons of crude were
spilled. There were no injuries.
• On February 15, 2017: A 36-inch-diameter Kinder Morgan natural gas
pipeline exploded and burned in Refugio County, Texas. There were no
injuries. The flames were visible 50 miles away. Refugio County Chief
Deputy Sheriff Gary Wright said the explosion occurred at an apparent
weak point in the pipeline that must have required maintenance, but KM
disputed the issue. Residents as far as 60 miles away thought it was an
earthquake, while others described it as "a thunder roll that wouldn’t end.”
According to the PHMSA incident listing, "the incident was most likely
caused by some combination of stress factors on the pipeline." The
explosion and resulting fire cost $525,197 in property damage. The pipe
was installed in 1964.
• On February 27, 2017: A crude oil pipeline ruptured in Falls City, Texas,
spilling about 42,630 gallons of crude oil. The cause was from internal
corrosion.
• On July 13, 2017: A contractor doing maintenance on Magellan's
Longhorn Pipeline hit that pipeline, in Bastrop County, Texas. About
87,000 gallons of crude oil were spilled, resulting in evacuations of nearby
residents.
• On December 13, 2017: An Energy Transfer Partners gas pipeline
exploded and burned, in Burleson County, Texas. There were no injuries
reported.
While pipelines are still the safest model of oil and gas
transportation, there are risks. Here are examples of pipeline problems throughout
the USA. From 1994 through 2013, the U.S. had 745 serious incidents with gas
distribution, causing 278 fatalities and 1059 injuries, with $110,658,083 in
property damage. From 1994 through 2013, there were an additional 110 serious
incidents with gas transmission, resulting in 41 fatalities, 195 injuries, and
$448,900,333 in property damage. From 1994 through 2013, there were an
additional 941 serious incidents with gas all system types, resulting in 363
fatalities, 1392 injuries, and $823,970,000 in property damage. It is interesting to
Sullivan
84
note that most pipeline accidents are discovered by local residents, not the
companies that own the pipelines!
PIPELINE MISHAPS IN THE UNITED STATES
Here are a sample of USA mishaps that costs both companies and the
USA money as well as lives lost:
• September 9, 2010: A PG&E 30-inch natural gas line exploded in San
Bruno, California and killed 8 people. Eyewitnesses reported the initial
blast "had a wall of fire more than 1,000 feet high".
• July 25, 2010: Crude oil pipeline ruptures near Marshall, Michigan,
spilling over 840,000 gallons of oil into the Kalamazoo River.
• December 12, 2012: A 20-inch transmission line owned by NiSource Inc.,
parent of Columbia Gas, exploded, leveling 4 houses, between Sissonville
and Pocatalico in Kanawha County, West Virginia (WV). When it blew,
nobody at pipeline operator, Columbia Gas Transmission knew it. An 800'
section of I-77 was completely destroyed. "The fire melted the interstate
and it looked like lava, just boiling." Later the West Virginia Public
Service Commission released several pages of violations by Columbia
Gas. Forty families were "impacted" by the explosion. The investigation
cited "external corrosion" as the cause of the blast.
• May 26, 2014: Viking gas pipeline explosion near Warren, Minnesota shot
a fireball over 100 feet in the air. Roads within a two-mile radius were
blocked off. Authorities suspected natural causes because there was still
frost in the ground and the soil was wet.
• November 16, 2017: TransCanada's Keystone Pipeline leaks 210,000
gallons (5,000 barrels) of crude oil in Marshall County (northeastern South
Dakota). Officials don't believe the leak affected any surface water bodies
or threatened any drinking water systems.
WORLDWIDE MISHAPS
While safety standards are better in the United States, both lives and
money has been lost with the pipeline problems. Since most USA oil companies
are multinational in nature, it is important to discuss the pipeline safety issues
with a global emphasis. Here are some examples of worldwide accidents that cost
both the companies and the countries lives and money:
• 1965: An explosion from a gas line destroyed several apartments in the
LaSalle Heights Disaster (LaSalle, Quebec), killing 28 people, the worst
pipeline disaster in Canadian history.
• 1978: A gas pipeline exploded and burned, killing 52 people in Colonia
Benito Juarez, Mexico, and injuring 11 in a town of only 100 people. The
failure created a crater 300 feet wide and 20 feet deep.
• 2006: A vandalized oil pipeline exploded in Lagos, Nigeria. Up to 500
people may have been killed.
Journal of Business and Accounting
85
• 2011: Nairobi pipeline fire kills approximately 100 people and
hospitalized 120.
• 2013: A Sinopec Corp. oil pipeline exploded in Huangdao, China. 55
people were killed.
• 2014: A pipeline blast in Southern Indian state of Andhra Pradesh killed
22 people and injured 37. The pipeline was operated by Gas Authority of
India Limited (GAIL) and a preliminary investigation revealed faulty
operational procedures as a cause of fire.
• 2014: A string of explosions originating in buried gas pipes occurred in
the city of Kaohsiung, Taiwan. Leaking gas, suspected to be propylene,
filled the storm drains along several major thoroughfares and the resulting
explosions turned several kilometers of road surface into deep trenches,
sending vehicles, people and debris high into the air and igniting fires over
a large area. At least 30 people were killed and over 300 injured.
CONCLUDING REMARKS
Pipelines are still the safest mode of transportation for oil and gas
products. People thinking about pipeline safety need to consider about the Exxon
Valdez accident and other tanker transport accidents. However, safety is of
utmost importance when conducting oil and gas business with the pipelines as the
primary mode of transportation. Please remember the differential in prices when
the product is an “Asset” and when the product is a “Liability”!
REFERENCES
Dey, P. K., Ogunlana, S. O., & Naksukakul, S. (2004). Risk-based maintenance
model for offshore oil and gas pipelines: A case study. Journal of
Quality in Maintenance Engineering, 10(3), 169-183.
Glukhova, M. G., & Zubarev, A. A. (2017). Economic appraisal of the program
of diagnostics of main gas pipelines, International Journal of Energy
Economics and Policy, 7(2)
Hellmann, H. (2015). Acknowledging the threat securing United States Pipeline
SACAD Systems, Energy Law Journal, 36(1), 157-178.
Jamal, F. (2015). Legal aspects of transnational energy pipelines: A critical
appraisal. European Networks Law and Regulation Quarterly (ENLR),
3(2), 103-116.
Klass, A. B., & Meinhardt, D. (2015). Transporting oil and gas: U.S.
infrastructure challenges. Iowa Law Review, 100(3), 947-1053.
Kolb, R. W. (2012). The natural gas revolution and central asia. The
Journal of Social, Political, and Economic Studies, 37(2), 141-180.
Leitzinger, J., & Collette, M. (2002). A retrospective look at wholesale gas:
Industry restructuring. Journal of Regulatory Economics, 21(1), 79.
Sullivan
86
Makholm, J. D., PhD., & Olive, L. T. W., PhD. (2016). The Politics of
U.S. Oil Pipelines: The First-Born Struggles to learn from the clever
younger sibling. Energy Law Journal, 37(2), 409-427.
Mbara, T., & Van, B. (2011). An analysis of perceptions and awareness of risk
associated with energy pipelines in South Africa. Journal of Transport
and Supply Chain Management, 5(1)
Mogel, W. A., & Gregg, J. P. (2004). Appropriateness of imposing common
carrier status on interstate natural gas pipelines, Energy Law Journal,
25(1), 21-56.
North America and Russia to dominate global oil and gas pipeline construction
by 2018. (2015). (). London: Global Data Ltd. Retrieved from
ABI/INFORM Global Retrieved from http://ezproxy.utpb.edu/login
Retrieved from ABI/INFORM Global.
Rigney, M. E. (2015). Clogging the pipeline: Exploring the D.C. circuit improper
segmentation analysis in Delaware riverkeeper network v. FERC and its
implications for the United States’ domestic natural gas production,
American University Law Review, 64(6), 1465-1502.
Journal of Business and Accounting
Volume 12, No 1; Fall 2019
87
AN ANALYSIS OF ALTERNATIVE WORK
ARRANGEMENTS TO ADDRESS THE GENDER GAP
IN PUBLIC ACCOUNTING
John Anderson
Hannah Smith San Diego State University
ABSTRACT Women make up approximately half of the full time staff at U.S.
accounting firms and one fourth of the partners and principals. These firms have
conducted extensive research to document the need for closing the gender gap
worldwide. Accounting firms are currently advocating alternative work
arrangements (AWAs) as part of a solution to close this gender gap. The purpose
of this study is to review the current advocacy for alternative work arrangements,
followed by a review of the academic literature, both theoretical and empirical, to
document the historical progress of and lessons learned in use and implementation
of AWAs. In this literature review, the paper specifically considers who are the
participants of AWAs and the outcomes to the employee and to the employer of
utilizing and/or implementing AWAs.
Keywords: Gender gap, public accounting, alternative work
arrangements, biases
INTRODUCTION Current business news articles document the lack of women in executive
positions. Fewer Fortune 500 chief executives are women than are named John
(Miller, Quealy, and Sanger-Katz, 2018). Women hold almost 52% of
professional-level jobs, earn 60% of master’s degrees, and represent 49% of the
college educated labor force (Herd and Jasik, 2018, p. 45). However, women still
appear to be hitting a ceiling at middle management. In 2017, 6.4% of the CEOs
at Fortune 500 companies were women (up from 4.8% in 2014), and women held
only 18% of board seats at S&P 1500 companies (Herd and Jasik, 2018, p. 45).
In March 2018 PricewaterhouseCoopers (PwC), published the results of
their international research report, concluding that the gender gap remains wide
and underemployment of women remains a pressing issue (PwC, 2018b, p.5). The
women in the survey were aged 28 to 40 and were at the point in their working
lives where the gap between men and women’s progression begins to widen
dramatically. This is also the time in women’s lives where the challenges of
combining careers and personal priorities increase (PwC, 2018a, p. 5).
According to the PwC report, the evidence suggests that both structural
and policy factors help explain the gender pay gap. The quantitative theory of
human capital investments suggests that since many women take time off to have
children (i.e., maternity leave) or plan to at some point take time off to have
Anderson and Smith
88
children, women may on average spend less time at work or have less incentive to
put in more hours, and therefore build up lower levels of human capital, which
could result in lower earnings (Erosa, Fuster, and Restuccia, 2016). Evidence also
supporting the quantitative theory of human capital investments was found in
analysis of MBA students (Bertrand, Goldin, and Katz, 2010).
The Peterson Institute showed in a 2015 study that companies with more
women at the executive and board levels perform better (Herd and Jasik, 2018).
Additionally, the academic literature finds that firms with women on boards and
in senior management have higher earnings quality and higher earnings
management restraint (Krishnan and Parsons, 2008; Srinidhi, Gul, and Tsui, 2011;
Arun, Almahrog, and Aribi, 2015; Kyaw, Olugbode, and Ptracci, 2015; Khliff and
Achek 2017; Duong and Evans 2016). The literature also finds that firms with
women on their boards have lower likelihood of fraud and internal control
weaknesses (Capezio and Mavisakalyan, 2016; Chen, Eshleman, and Soileau,
2016; Khliff and Achek, 2017).
To close the gender gap in positions of leadership, current literature from
all four of the largest public accounting firms currently advocates the need for
alternative work arrangements (AWAs) and a work environment where women
can thrive (Deloitte, 2018; Ernst and Young, 2017; KPMG, 2018; PwC, 2018b).
Given this emphasis in current practitioner literature on the need for
AWAs and the potential AWAs have to influence positively the gender gap in
public accounting, the following academic literature review will focus on AWAs.
This review of the academic literature on AWAs starts by examining (1) who
participates in the AWAs, followed by a discussion of (2) the outcomes to
employees who participate in AWAs, and concluding with a discussion of (3) the
outcomes to employers who offer AWAs.
LITERATURE REVIEW Who Participates in AWAs: Non-Accounting Specific Literature
First, the literature that examines who participates in AWAs in a non-
accounting related workplace is discussed, followed by a discussion of literature
that examines AWA participation in a specifically accounting related context. In
examining who participates in AWAs, Hall (1990) finds that both men and women
face struggles in balancing a family and a career. Specifically, he speculates that
men and women with children especially face the struggles of balancing family
and career thus suggesting that both men and women with children face the
necessity of participating in AWAs. Additionally, he finds that some employees,
specifically those with children, may be participating in their own form of AWAs
even when formal AWAs are not provided. Informal AWAs can manifest through
an employee arriving outside the traditional start time, working through lunch, or
taking some work home (Hall 1990).
A unique finding in Kossek, Noe, and DeMarr (1999) is that firms that
have formal AWAs may not actually be the firms that have employees working in
AWAs. They suggest that it takes more for a firm than simply to adopt AWA
policies to have employees utilizing AWAs. Subsequent research shows that
Journal of Business and Accounting
89
organizational culture supportive of AWAs is required for employees to actually
participate in and use AWAs (Kossek, Barber, and Winters 1999).
Kossek, et al. (1999a), using a field study design, examine one facet of
organizational culture: peers’ influence on an employee utilizing an AWA. They
find that employees’ with peers that previously utilized (or planned to utilize in the
future) AWAs were significantly more likely to actually utilize an AWA
themselves compared to those without such peers. They also found that women
were more likely to utilize AWAs than men were. They specifically examined
flexible schedules, part-time work, and leaves of absence. Their results suggest
that organizational culture strongly drives whether an individual participates in an
AWA. Besides the organizational culture aspect of peer influence, specific job
characteristics influence whether an individual participates in an AWA (Powell
and Mainiero, 1999).
After an organization adopts policies allowing for formal AWAs,
individuals interested in these arrangements must request and obtain approval by
someone at the firm for this arrangement (Powell and Mainiero, 1999). Work
disruption theory suggests that managers consider how much work the
arrangement will disrupt if they approve an arrangement (Powell and Mainiero,
1999). Specifically, AWAs are more likely to get approved for and thus get used
by those working on less critical tasks or having fewer special skills, those without
supervisory roles, and those that suggest a more temporary arrangement (Powell
and Mainiero, 1999). While not everyone may be allowed an AWA, work/family
border theory suggests that some sort of work/life integration is occurring whether
or not formal programs are used (Clark 2000).
Work/family border theory developed in Clark (2000) suggests that
everyone with an outside-the-home job must figure out how to integrate and
balance their outside work with their family life. With work family
balance/integration as the goal, Clark’s border theory predicts as follows: (1)
AWAs will benefit those where work and family are more similar, (2) AWAs will
benefit individuals that prioritize family over work, (3) individuals with more
influence and control both at home and work will benefit more than those with less
influence and control in either home or work. AWAs according to this theory are
more likely to benefit individuals and situations with certain characteristics.
However, this theory also suggests that regardless of whether a formal AWA
exists, work life integration may occur with non-ideal balancing of work and life
demands.
Guest (2002) discusses why work-life balance is currently an important
issue. “[T]he demands of work contribute to a reduced participation in non-work
activities resulting in an imbalance.” This relates to AWAs as most view AWAs
as promoting or increasing work-life balance. Work-life balance has become a
bigger issue with women in the workforce, since the intensity of work as well as
work hours has been increasing over time, and since Generation X (a group that
values a good work life balance) is joining the workforce. Whether work and life
are in balance depend on a number of factors including workplace and job
characteristics, family culture and family demands, and individual characteristics
Anderson and Smith
90
such as work orientation, personality, energy, gender, age, career stage, etc. (Guest
2002). While work/life balance depends on certain factors and characteristics, use
of family-friendly benefits such as AWAs also relates to certain factors and
characteristics, but not in all of the expected ways (Butler, Gasser, and Smart,
2004). Thus, the necessity for AWA use has increased as more women have
entered the workforce over the decades.
Butler, et al. (2004) examine individual psychological characteristics as
well as demographics that predict participating in and using family-friendly
benefits (including AWAs). They find that gender and marital status do not
influence, while age of youngest child does influence the likelihood of using
family-friendly benefits. They also find that work outcome expectancies, but not
family outcome expectancies influence an individual’s decision to use the family
friendly benefit. While these are interesting and potentially useful findings, the
generalizability of the results found in that study may be limited due to the sample
used (i.e., the parents and family of college students at a university) (Butler, et al.
2004).
Who Participates in AWAs: Specific Accounting Literature
Almer, Cohen, and Single (2003) examine who is interested in and who
actually participates in AWAs in the assurance line of public accounting firms.
They examine individual and firm characteristics and find that an employee is
more likely to express interest in adopting an AWA if he or she perceives the
existence of organizational support and support from those who review him or her.
Additionally, they find that family considerations, gender, marital status, and
position at the firm all impact an employee’s likelihood of participating in an AWA
(Almer et al., 2003).
Gallhofer, Paisey, Roberts, and Tarbert (2011) examine the work-lifestyle
choices of female accountants in Scotland, with special focus on whether women
face employer imposed constraints or self-made constraints via choices related to
prioritizing family. Using surveys and interviews to gather data, they find that
women face some organizational issues in the workplace, but often choose family
over work thus leading to lost work opportunities. One way in which they choose
family over work is by choosing to use AWAs. Further, they find that women
generally prefer to forego their lost opportunities in order to prioritize family. The
findings support preference theory that suggests, “women’s choices owe less to
inequalities in the workplace and more to the preferences of individual,
particularly, but not exclusively, women”. The findings also support difference
feminism, which allows women and men to operate differently due to inherent
differences and choices. Preference theory and difference feminism, both
supported here, suggest that women are more likely to choose AWAs to allow for
more family time (Gallhofer et al., 2011).
Buchheit, Dalton, Harp, and Collingsworth (2016) examine multiple types
of organizations including large public accounting firms, smaller public
accounting firms, and industry accounting departments. They find that employees
Journal of Business and Accounting
91
working at Big 4 firms are least likely of all the groups surveyed to feel that they
could actually utilize an AWA. Moreover, they find no difference between those
working in audit or tax lines of service. In addition, they find that women feel more
organizational support for AWA use than men do, which supports the traditional
gender role theory found in Johnson, Lowe, and Reckers’s (2008) study.
Further, most accounting firms now offer AWAs. The accounting firms
specifically design their AWAs to attract female accountants. In addition, these
marketing efforts around AWAs have been successful with women being much
more likely to actually utilize AWAs than are men (Cohen, Dalton, Holder-Webb,
and McMillan, 2018; Almer et al., 2003; Dalton, Cohen, Harp, and McMillan,
2014; Johnson et al. 2008). However, some question whether these AWAs are
usable or underutilized (Dambrin and Lambert 2008). In interviews with French
Big Four employees, Dambrin and Lambert (2008) find that women are perceived
as utilizing AWAs at the time right around when they have their first child.
Additionally, those interviewed thought those working in taxation may be more
able to utilize AWAs. Lupu (2012) also interviewing French accountants finds
that women with or about to have children are the ones seeking AWAs. In addition,
senior women who have proven themselves in their jobs prior to seeking an AWA
are more likely to be the ones approved to work an AWA. Whiting and Wright
(2001) in an examination of the New Zealand accounting profession also find that
women around the time they become mothers are the ones seeking out AWAs.
In an analysis of one Big 4 accounting office, Kornberger, Carter, and
Ross-Smith (2010) found that in 2009, 245 females and 40 males participated in
the firm’s AWA program. They also found evidence that women raising a family
were the ones most utilizing AWAs, which supports prior literature (e.g., Dambrin
and Lambert 2008, Whiting and Wright 2001).
Outcomes to Employee from AWAs: Non-Accounting Specific
Literature
Next, the outcomes to the employee from working an AWA are examined,
focusing first on research in non-specific accounting literature, followed by a
discussion of specifically accounting literature. The literature finds that the main
benefits of AWAs to the employee participating in the AWA are reduced stress
and burnout as well as increased job satisfaction and commitment (Johnson, et al.,
2008; Kossek and Ozeki, 1999; Baltes et al., 1999; Powell and Mainiero, 1999,
Rhodes, 2001; Scandura and Lankau, 1997). The main drawbacks of AWAs to the
employee participant in the AWA are reduced promotions, raises, and status within
the organization and assignment to less challenging work (Johnson et al., 2008;
Butler et al., 2004; Judiesch and Lyness, 1999; Nord, Fox, Phoenix, and Viano,
2002; Rogier and Padgett, 2004; Butler et al., 2004; Hall, 1990; Kossek et al.,
1999a).
Justice theory suggests that workers believe rewards should be allocated
in a certain way, specifically, equity based allocation should be used when
productivity is the goal, equality based allocation should be used when team
Anderson and Smith
92
building and good social relationships are the goals, and need based should be used
when social responsibility are the goals. According to justice theory, at for-profit
firms, profitability is the goal and thus managers should allocate rewards based on
productivity. Providing family friendly policies that benefit only those with
children provides rewards based on a need basis, which would be appropriate when
social responsibility is the goal, but not when profitability is the goal. Thus,
resentment by others may result when an employee utilizes an AWA. (Rothausen,
Gonzalez, Clarke, and O’Dell, 1998)
Hill, Miller, Weiner, and Colihan (1998) examine one specific type of
AWA, telecommuting, and examine its effects on work-life balance. Interestingly,
they find that employees perceive telecommuters to have greater productivity,
higher morale, more flexibility, longer working hours, and less teamwork than
non-telecommuters; however, when they examine actual outcomes, they find that
productivity and flexibility only are improved. This paper’s most interesting
finding is that what is perceived by the employees is not the same as what actually
occurs. This finding may explain other mixed findings in the literature when a
study relies on perceptions instead of actual effects (Hill et al., 1998).
Judiesch and Lyness (1999), using a field study design, examined the
impact of leaves of absence on career promotions and salary increases. In line with
what both human capital theory and gendered culture theory predict, they found
that employees taking a leave of absence faced fewer subsequent promotions and
smaller salary increases. Human capital theory says that leaves of absence prevent
at least some human capital from accruing, thus those taking a leave of absence
have lower human capital. Additionally, gendered culture theory says that
organizational cultures reflect male values and, a leave of absence, as it is often
associated with maternity leave, goes against the organizational culture. (Judiesch
and Lyness, 1999).
Studies have shown that fewer employees use AWAs than was expected
when the firms implemented the AWAs (Schwartz, 1994). As discussed
previously, it takes more than just the offering of AWAs for employees to utilize
AWAs (Kossek et al., 1999a). Schwartz (1994) has shown that employees avoid
fully utilizing these arrangements due to fears of negative career consequences
including delayed career advancement and not feeling truly free to take on and
adopt available AWAs. This research has also shown that these fears are somewhat
true and that those taking on AWAs do experience career penalties related to their
work arrangements (Schwartz, 1994).
Clark (2001) examines the flexibility of working hours and the flexibility
of work itself as it relates to work/family balance for individuals likely to be facing
difficulty balancing work/family. This study focuses on the outcomes to the
employee at home and at work. She finds that more flexible work increased
satisfaction and well-being with work and family, while flexibility of work time
did not have any benefits. In a univariate test, they do find results with flexibility
of work time, but after controlling for flexibility of the work itself, the flexibility
of work time no longer explains anything (Clark 2001).
Journal of Business and Accounting
93
In a qualitative study on the implications of a specific type of AWA,
reduced workload, Lee et al. (2002) found that the employees participating in the
AWA were generally quite happy and satisfied with their arrangement and the
potential tradeoffs they were facing. Specifically, the participants seemed to think
that their AWA participation affected their short-term career advancement.
Further, the researchers found that a combination of an employee’s individual
characteristics as well as organizational and job specific characteristics influenced
the success of the AWA. Specifically, the individual characteristics that led to a
successful arrangement included hard working, self-motivated, having unique
skills, and organized. The job specific and organizational characteristics that
seemed to have a positive effect on AWA were having project-oriented work,
supportive coworkers, and an organizational culture that promoted employee-
friendly values. The most important characteristic they found was having a
supportive boss. This study was unique in that it focused on high-level managers
and professionals, which are the types of employees and jobs that most perceive as
being less suited to an AWA (Lee et al., 2002).
A study on another type of family friendly policy, specifically leaves of
absence, found that men faced a social penalty when taking a leave, whereas
women did not (Wayne and Cordeiro, 2003). Specifically, in an experimental
setting, participants rated men as less likely to be good organizational citizens at
work when choosing to take a leave of absence. Organizational citizenship
behaviors are extra behaviors or acts that an employee performs that are beyond
the job description (Wayne and Cordeiro, 2003). While organizational citizenship
is clearly beneficial to the employer, the employee does not go unrewarded. In fact,
research has found that employees perceived to be better organizational citizens
receive higher performance ratings and rewards (MacKenzie, Podsakoff, and
Fetter 1991; Borman and Motowidlo, 1997; Wayne and Cordeiro, 2003).
Butler and Skattebo (2004) examine the performance rating of employees
after experiencing a family-work conflict. They find the performance rating for
men declines while the performance rating for women is unaffected. Eagly’s
(1987) social role theory and Dipboye’s (1985) stereotype fit model predict the
results the authors find in this study as the behavior is inconsistent with traditional
gender roles. Interestingly, the gender of the evaluator does not affect the results –
i.e., both genders seem to punish men for deviating from the traditional gender
roles (Butler and Skattebo, 2004).
Rogier and Padgett (2004) examine the perceptions of career advancement
for women who participate in AWAs in an experimental setting. In their study,
they find that participants perceive women as having less career advancement
potential, specifically lower job and career dedication and lower motivation for
advancement. These results are consistent with other research findings that women
participating in AWAs face negative career advancement perceptions.
Smithson and Stokoe (2005) use focus groups to examine the participants’
perceived repercussions from participation in AWAs. They find that the
participants perceive those participating in AWAs as being less committed and as
choosing family more over career. Further, they find that resentment arises against
Anderson and Smith
94
those that are participating, as the idea exists that someone else will need to pick
up the work from the employee participating in the AWA.
Outcomes to Employee from AWAs: Specific Accounting Literature
Next, the outcomes to the employee from AWA participation in the
accounting industry specifically are examined. In one study (Almer and Kaplan,
2002), the authors surveyed a group of CPAs working in public accounting. They
compared those that were currently working in an AWA with those that were likely
candidates for working in an AWA. The study focused on characteristics relating
to the employee’s feelings about their job – i.e., job satisfaction, turnover intention,
burnout, and stress. They found that employees in an AWA had greater job
satisfaction, reduced plans of turnover, less burnout, and less stress as compared
to those on a traditional work schedule. Further, the survey found these same
results occurred when employees switched from a traditional schedule to an AWA.
In the firms studied, the employee generally had to have a conversation with his or
her superiors and define their workload, assignments, working arrangement, etc.
to begin an AWA. The authors suggest that this conversation may have an impact
on the employee’s feelings about his or job. The authors surmise that this
conversation and role defining may be an unintended benefit of AWA participation
(Almer and Kaplan, 2002).
While Almer and Kaplan (2002) find benefits to participation in AWAs,
they do not examine potential negative consequences. In a research design where
the participants evaluate their “peers” at a public accounting firm, Cohen and
Single (2001) find that participants rated those engaged in an AWA as having
lower likelihood of professional success, higher anticipated turnover, and less
likely to be chosen on a subsequent engagement than those not working in an
AWA.
Frank and Lowe (2003) examine participation in AWAs in private
accounting. They examine the impact participation in AWAs has on performance
evaluations, job commitment, and career progression and whether gender
influences any effect. They find that participation in an AWA did not affect short-
term current task performance or job commitment and dedication but did
negatively influence perceptions of long-term career potential. Specifically, they
found that those in an AWA were perceived as more likely to perform poorly in
the future, less likely to be chosen for special projects, more likely to be given less
challenging work, and less likely to be promoted. In addition, they find that gender
of the individual participating in the AWA did not seem to influence the results.
Further, they find evidence that those that are currently engaged in an AWA do not
have negative perceptions of long-term career prospects (Frank and Lowe 2003).
These findings from private accounting contribute to the other papers discussed in
this section and suggest that the negative actual and negative perceptions
surrounding those who participate in AWAs may extend beyond traditional public
accounting work. Additionally, some of these same negative consequences of
Journal of Business and Accounting
95
AWA participation researchers also found in non-accounting industry jobs as
discussed previously (e.g., Rothausen et al., 1998, Schwartz 1994).
Johnson et al. (2008), using manager level employees at Big 4 accounting
firms, examine the effect of AWA use on potential career advancement in an
experimental setting. Using unique questioning, they are able to disentangle the
effects AWA participation has on formal evaluation separate from informal
evaluations. Consistent with firms promoting their work-life balance initiatives,
they find that AWAs do not impact formal evaluations; however, participation in
AWAs does appear to influence informal evaluations. Further, consistent with
traditional organizational role views, they find that participants informally rated
males participating in AWAs lower than females participating in AWAs (Johnson
et al., 2008).
Further, organizational policies and decisions, including providing AWAs
at a firm that attempt to help women achieve work-life balance may in fact be
harming women’s careers (Lyness and Thompson, 1977; Cohen et al., 2018). The
mechanism may be that women are assigned lower risk jobs that also have fewer
developmental opportunities in order to help women achieve work-life balance
(Lyness and Thompson, 1977, Cohen et al., 2018). Findings from interviews done
with French Big 4 employees further support this, where they find that employees
perceive those that utilize AWAs will face marginalization on the path to
partnership. Those interviewed suggested that those utilizing AWAs were likely to
be put on a different path not leading to partnership, such as director track
(Dambrin and Lambert, 2008). Additionally, those utilizing AWAs may face a
“drop in status” and significantly lower salary (Dambrin and Lambert, 2008).
Dalton et al. (2014) find that employees feel lower levels of gender
discrimination at firms that are more supportive of AWAs. However, in line with
prior research (e.g., Kossek, et al., 1999b), Dalton et al. (2014) find that the benefit
of decreased gender discrimination at a firm requires actual organizational support
for AWAs and not just the offering of the AWA. Further, they find that the
decreased gender discrimination at a firm from support for AWAs also leads to
lower levels of turnover intention. This suggests that individuals at firms with
support for AWAs experience lower levels of gender discrimination, which leads
to them staying at their job and not searching for new work, which may or may not
be a positive outcome to the employee. This finding of lower turnover when AWAs
and flexibility in work are offered is also in line with what Dambrin and Lambert
(2008) find in their interviews of employees at French Big Four firms. Specifically,
they find that not offering flexibility is perceived to lead to higher turnover.
In structured interviews with employees at a Big 4 accounting firm office,
Kornberger et al. (2010) find evidence that employees participating in AWAs face
a questioning of their individual work performance due to less “face time” in the
office during normal office hours. Additionally, these AWA participants faced
“greater managerial surveillance when they were at work”, not being selected for
the “large and prestigious engagements”, the perception of not being able to
appropriately serve clients, and the perception of “not being serious” about work.
They further found that this lower perceived visibility due to AWA usage led to
Anderson and Smith
96
negative perceptions of career progression within the firm. Lupu (2012) finds
similar results in a French setting. Specifically, the author finds that those working
AWAs face organizational marginalization through exclusion from important
client meetings, a path on a non-partner track, the perception of partial
disengagement, lack of loyalty, and incompatibility for a management position.
Additionally, in a large scale analysis of women in accounting literature, Haynes
(2017) comes to the conclusion that women utilizing AWAs face stereotyping,
lower pay, gendered hierarchies, and discrimination.
Public accounting firms in the U.S. have adopted AWAs to counteract the
gender gap that results from women leaving public accounting before moving up
in the firm. However, if participating in these AWAs has negative consequences
to career advancement, these AWAs may not be serving their intended purpose.
Outcomes to Employer from AWAs: Non-Accounting Specific
Literature
Next, research that has examined outcomes to the employer from
implementing or having AWAs is discussed. First, this literature is examined in
the context of any type of business, and then examined by looking solely at what
research has found in the accounting industry related to the benefits and drawbacks
on employers who offer AWAs.
Prior literature finds that the main benefits to the employer of offering and
having their employees participate in an AWA are increased productivity and
reduced turnover (Johnson et al., 2008; Hill et al., 1998; Nord, et al., 2002; Connor,
Hooks, and McGuire, 1999; Dalton and Mesch, 1990). The main drawbacks to the
employer of offering and having their employees participate in an AWA are
increased difficulty of work scheduling and job coverage, potential declines in
quality or client service, increased supervisor workload, and resentment by non
participants (Johnson et al., 2008; Nord et al., 2002; Baltes et al., 1999; Kossek
and Ozeki, 1999; Lee et al., 2002; Parker and Allen, 2001).
Flexibility in the workplace can help to create balance between home and
work life (Hall 1990). This flexibility does not have to always be a formal AWA.
Hall’s (1990) work suggests that even firms that do not have formal AWAs in place
likely already have informal AWAs in place, and thus formalizing AWAs may
have potential benefits – i.e., retaining top employees.
In a naturally occurring field experiment related to the implementation of
an AWA, employee absenteeism decreased significantly during the period of
flexible work arrangements, but employee turnover was unaffected (Dalton and
Mesch, 1990).
Scandura and Lankau (1997) examine how the perceptions of an
organization having AWAs affected organizational commitment and job
satisfaction of the employees. Research has found job satisfaction is a prerequisite
for organizational commitment by the employee (Scandura and Lankau, 1997;
Vandenberg and Lance, 1992, Williams and Hazer, 1986). In addition,
organizational commitment is associated with and may lead to many positive
Journal of Business and Accounting
97
benefits to the employer including beneficially affecting: performance, turnover,
participation, power, teamwork, and professionalism (Scandura and Lankau, 1997;
Aranya, Kushnir, and Valency, 1986; Mathieu and Zajac, 1990; Welsch and
LaVan, 1981). Scandura and Lankau (1997) find that employees working at
organizations where they perceive AWAs exist (regardless of whether they
themselves participate) self-report higher levels of job satisfaction and
organizational commitment. Further, these results are stronger for women and
those with children.
Since companies are in the business of making a profit, for an employer to
set up an AWA there must be some benefit to the employer. Baltes et al. (1999)
find that the effects of AWAs to the employer are positive overall with the
employees working in an AWA having improved productivity/performance, job
satisfaction, absenteeism, and satisfaction with work schedule. These outcomes
additionally benefit the employer and are not short lived but rather are long term.
However, these findings only apply to lower level employees (Baltes et al., 1999).
Employees often underutilize family friendly policies, including AWAs,
provided by the employer (Kossek and Ozeki, 1999; Kossek et al., 1999b). Some
researchers also found these policies to be ineffective in achieving the goals
thought to be important to the employer (Kossek and Ozeki, 1999). According to
Kossek and Ozeki (1999), mixed evidence at best is found regarding AWAs benefit
to the employer in the form of lower turnover, lower absenteeism, and increased
employee commitment. When measuring turnover intentions, some research finds
no effect, but some research finds lower turnover when AWAs are put into place
(Dalton and Mesch, 1990; Dunham, Pierce, and Castenada, 1987; Pierce and
Newstrom, 1982; Rothausen, 1994). Regarding absenteeism, research found that
certain specific types of AWAs worked in some situations and not in others (Dalton
and Mesch, 1990; Krausz and Freibach, 1983; McGuire and Liro, 1987; Pierce and
Newstrom 1982, 1983; Thomas and Ganster, 1995). Thus, the benefits, if any,
accruing to the employer seem to be very specific to the type of AWA and the type
of workplace examined.
A major detriment and hurdle to overcome in achieving successful AWAs
is feelings of resentment and unfairness towards those participating in AWAs felt
by the non-participants (Rothausen et al., 1998). In a study examining individual
employee characteristics related to perceptions of unfairness, the authors found
that “younger workers, minorities, those who had used flexible work arrangements,
and workers in jobs requiring a greater degree of task interdependence had more
favorable perceptions concerning work/family benefits than did older workers,
Caucasians, individuals who had not used flexible work arrangements, and those
working in jobs requiring a lesser degree of task interdependence” (Parker and
Allen, 2001). Hegtvedt, Clay-Warner, and Ferrigno (2002) also study potential
resentment by those employees that are not able to take advantage of the family
friendly benefits, including AWAs. Those that do not utilize AWAs may feel
resentful because they are not getting the benefit, and think their employer may
expect them to pick up extra work. Using the National Survey of the Changing
Workforce, Hegtvedt et al., (2002) find that 40 percent of the employees in the
Anderson and Smith
98
survey would feel resentment if their employer provided family friendly benefits
that did not help them personally. They also found that 16 percent of those
surveyed would feel resentment if they had to pick up extra work due to the family
and personal commitments of others. They find that a supportive workplace
reduces resentment. Additionally, they find women feel less resentment than men
in regards to picking up extra work due to someone else’s family commitments,
but both men and women feel the same level of resentment towards family friendly
benefits that do not personally benefit them (Hegtvedt et al., 2002).
One of the greatest benefits to the employer of offering AWAs is in
recruiting and potentially retaining workers who value this benefit (Nord et al.,
2002). Besides recruiting and retaining top employees, most of the benefits of
AWAs seem to accrue to the employee participating (Nord et al., 2002). Some of
the indirect costs of these programs to the employer includes loss of productivity
and feelings of unfairness by non-participants (Nord et al., 2002).
As some studies have found that women more often push for and then
utilize AWAs than men (Goodstein, Gautam, and Boeker, 1994; Powell and
Mainiero, 1999), and AWAs have been shown in some studies to increase
productivity and reduce turnover (Johnson et al., 2008; Hill et al., 1998; Nord et
al., 2002; Connor et al., 1999; Dalton and Mesch, 1990), it stands to reason that
firms with AWAs find particular value in the skills women bring to their
organizations. Parker (2008) examined differences in strategic management and
accounting processes caused by gender differences. He finds evidence that both
feminine and masculine features are important in the workplace (Parker, 2008).
Women tend to use more “interactive, participative leadership styles and to exhibit
the transformational leadership style in particular” while men tend to use a
“transactional leadership” style (Parker, 2008; Burke and Collins, 2001; Kim and
Shim, 2003; Trinidad and Normore, 2005; Wilson, 1995). Further research has
found that women make more “ethically based decisions than men” (Glover, 2002;
Parker, 2008).
Outcomes to Employer from AWAs: Specific Accounting Literature
Implementing AWAs at accounting firms and in the accounting industry
specifically is extremely important as Cohen and Single (2001) found out while
interviewing Human Resources and Auditing personnel at a public accounting firm
for a case study, and as is also evidenced through the literature the large accounting
firms publish related to their AWA availability (e.g., PwC, 2018b, p. 6). In the
Cohen and Single (2001) study, participants thought women were much more
likely to be interested in AWAs than men were. Losing a manager and retraining
someone to take his or her place costs a firm approximately 150% of the manager’s
salary (Coolidge and D’Angelo, 1994). This suggests potential benefit from
retaining an employee even if an AWA has significant costs. Additional potential
consequences to accounting firms of offering AWAs include: (1) the inability of
the individual on a flexible work arrangement to stay career focused and generate
new business, (2) increased commitment to the firm by the individual, and (3) once
Journal of Business and Accounting
99
a program is put into place, doing away with it would generate negative publicity
(e.g., Yahoo CEO Marissa Mayer’s 2013 ban on working remotely).
Additionally, in interviews with French Big 4 employees, interviewers
found evidence of resentment and feelings of injustice towards those participating
in AWAs, which follows what is found in non-accounting industries (Dambrin and
Lambert 2008; Rothausen et al., 1998; Hegtvedt et al., 2002). Also in a French
setting, Lupu (2012) finds that managers have more difficulty managing
employees who participate in AWAs than employees who follow normal working
schedules.
The perception that AWAs may prevent fully servicing a client’s need is
a negative outcome of AWA participation found in Kornberger et al.’s (2010)
study. However, they also found a positive outcome in providing AWAs at a Big
4 accounting firm: AWAs provided benefits to the reputation of the firm in the
professional service and general business community. They find that the AWA
program actually enhanced gender barriers at the firm as the formal AWA program
gave the firm an excuse as to why women were not progressing in the firm.
CONCLUSION
The gender pay gap and lack of women at top levels of
management is an ongoing issue in the United States and abroad in large public
accounting firms. In the long term, a comprehensive solution involving public
policy may be necessary. However, in the short term, accounting firms are trying
to utilize AWAs to attempt to narrow this gap. Effectively implementing AWAs,
prior literature finds, is possibly even more important than just the offering of
AWAs. AWAs may have unintended consequences for both the participant and the
offering firm. Careful analysis of the specifics of the AWA program as well as the
implementation of the AWA will be necessary to ensure the firms meet the
intended goals. However, the unintended costs of AWAs can be quite large.
Accounting firms will need to carefully identify their goals and continuously
measure their progress to be successful. By examining in this article the literature
that studies historical results of implementation and/or use of AWAs, this paper
sheds some light on how best to structure and implement AWAs.
REFERENCES
Almer, E., Cohen, J., and Single, L. (2003). Factors affecting the choice to
participate in flexible work arrangements. Auditing: A Journal of Practice
and Theory, 22, 69-91.
Almer, E. and Kaplan, S. (2002). The effects of flexible work arrangements on
stressors, burnout and behavioral job outcomes. Behavioral Research in
Accounting, 14, 1-34.
Aranya, N., Kushnir, T., and Valency, A. (1986). Organizational commitment in a
male-dominated profession. Human Relations, 39, 433-448.
Anderson and Smith
100
Arun, T.G., Almahrog, Y.E., and Aribi, A.Z. (2015). Female directors and earnings
management: evidence from UK companies. International Review of
Financial Analysis, 39, 137-146.
Baltes, B., Briggs. T., Huff, J., Wright, J., and Neuman, G. (1999). Flexible and
compressed workweek schedules: a meta-analysis of their effects on work-
related criteria. Journal of
Applied Psychology, 84, 496-513.
Bertrand, M., Goldin, C., and Katz, L. (2010). Dynamics of the gender gap for
young professionals in the financial and corporate sectors. American
Economic Journal: Applied Economics, 2, 228-255.
Borman, W. and Motowidlo, S. (1997). Task performance and contextual
performance: The meaning for personnel selection research. Human
Performance, 10, 99–109.
Buchheit, S., Dalton, D., Harp, N., and Hollingsworth, C. (2016). A contemporary
analysis of accounting professionals’ work-life balance. Accounting
Horizons, 30, 41-62.
Burke, S. and Collins, K.M. (2001). Gender differences in leadership styles and
management skills. Women in Management Review, 16, 244-56.
Butler. A., Gasser. M., and Smart, L. (2004). A social-cognitive perspective on
using family-friendly benefits. Journal of Vocational Behavior, 65, 57-70.
Butler, A. and Skattebo, A. (2004). What is acceptable for women may not be for
men: the effect of family conflicts with work on job-performance ratings.
Journal of Occupational and Organizational Psychology, 77, 553-564.
Capezio, A. and Mavisakalyan, A. (2016), Women in the boardroom and fraud:
evidence from Australia. Australian Journal of Management, 41, 719-734.
Chen, Y., Eshleman, J.D., and Soileau, J. (2016). Board gender diversity and
internal control weaknesses. Advances in Accounting, Incorporating
Advances in International Accounting, 33, 11-19.
Clark, S. (2000). Work/family border theory: a new theory of work/family balance.
Human Relations, 53, 747-770.
Clark, S. C. (2001). Work cultures and work/family balance. Journal of Vocational
Behavior, 58, 348-365.
Cohen, J., Dalton, D., Holder-Webb, L., and McMillan, J. (2018). An analysis of
glass ceiling perceptions in the accounting profession. Journal of Business
Ethics, 1-22.
Cohen, J. and Single, L. (2001). An examination of the perceived impact of flexible
work arrangements on professional opportunities in public accounting.
Journal of Business Ethics, 32, 317-328.
Connor, M., Hooks, K., and McGuire, T. (1999). Gaining legitimacy for flexible
arrangements and career paths: The business case for public accounting and
professional service firms. In S. Parasuraman & J. Greenhaus (Eds.),
Integrating work and family: Challenges and choices for a changing world.
Westport, CT: Praeger.
Coolidge, L. and D'Angelo, D. (1994). Family issues to shape the profession's
future. The CPA Journal, 16-21.
Journal of Business and Accounting
101
Dalton, D., Cohen, J., Harp, N., and McMillan, J. (2014). Antecedents and
consequences of perceived gender discrimination in the audit profession.
Auditing: A Journal of Practice & Theory, 33, 1-32.
Dalton, D. and Mesch, D. (1990). The impact of flexible scheduling on employee
attendance and turnover. Administrative Science Quarterly, 35, 370-387.
Dambrin, C. and Lambert, C. (2008). Mothering or auditing? The case of two big
four in France. Accounting, Auditing & Accountability Journal, 21, 474-506.
Deloitte. (2018). The gender dividend: Making the business case for investing in
women. Retrieved from
https://www2.deloitte.com/content/dam/Deloitte/ru/Documents/public-
sector/gender-dividend-en.pdf .
Dipboye, R. L. (1985). Some neglected variables in research on discrimination in
appraisals. Academy of Management Review, 10, 116–127.
Dunham, R.B., Pierce, J.L., and Castenada, M.B. (1987). Alternative work
schedules: Two field quasi-experiments. Personnel Psychology, 40, 215-
242.
Duong, L. and Evans, J. (2016). Gender differences in compensation and earnings
management: evidence from Australian CFOs. Pacific-Basin Finance
Journal, 40, 17-35.
Eagly, A. (1987). Sex differences in social behavior: A social-role interpretation.
Hillsdale, NJ: Erlbaum.
Ernst and Young. (2017). Five lessons from EY’s gender gap journey. Retrieved
from https://www.ey.com/en_gl/workforce/five-lessons-from-eys-gender-
gap-journey .
Erosa, A., Fuster, L., and Restuccia, D. (2016). A quantitative theory of the gender
gap in wages. European Economic Review, 85, 165-187.
Frank, K. and Lowe, D.J. (2003). An examination of alternative work
arrangements on perceptions of performance and career success in private
accounting practice. Accounting Horizons, 17, 139-152.
Gallhofer, S., Paisey, C., Roberts, C., and Tarbert, H. (2011). Preferences,
constraints and work-lifestyle choices: The case of female Scottish chartered
accountants. Accounting, Auditing & Accountability Journal, 24, 440-470.
Glover, S.H. (2002). Gender differences in ethical decision making. Women in
Management Review, 17, 217-27.
Goodstein, J., Gautam, K., and Boeker, W. (1994). The effects of board size and
diversity on strategic change. Strategic Management Journal, 15, 241-50.
Guest, D. (2002). Perspectives on the study of work-life balance. Social Science
Information, 41, 255-279.
Hall, D. (1990). Promoting work/family balance: an organization-change
approach. Organizational Dynamics, 18, 5-18.
Haynes, K. (2017). Accounting as gendering and gendered: a review of 26 years
of critical accounting research on gender. Critical Perspectives on
Accounting, 43, 110-124.
Anderson and Smith
102
Hegtvedt, K., Clay-Warner, J., and Ferrigno, E. (2002). Reactions to injustice:
factors affecting workers' resentment toward family—friendly policies.
Social Psychology Quarterly, 65, 386-400.
Herd, W.W. and Jasik, M. (2018). The Business Case (and plan) for Gender
Equality. The Future of Everything, A Look Ahead From the Wall Street
Journal. May 2018, 44.
Hill, E., Miller, B., Weiner, S., and Colihan, J. (1998). Influences of the virtual
office on aspects of work and work/life balance. Personnel Psychology, 51,
667-683.
Johnson, E., Lowe, J., and Reckers, P. (2008). Alternative work arrangements and
perceived career success: Current evidence from the big four firms in the
US. Accounting, Organizations and Society, 33, 48-72.
Judiesch, M. and Lyness, K. (1999). Left behind? The impact of leaves of absence
on managers' career success. Academy of Management Journal, 42, 641-
651.
Khliff, H. and Acheck, I. (2017). Gender in accounting research: a review.
Managerial Auditing Journal, 32, 627-655.
Kim, H.S. and Shim, S. (2003). Gender-based approach to the understanding of
leadership roles among retail managers. Human Resource Development
Quarterly, 14, 321-42.
Kornberger, M., Carter, C., and Ross-Smith, A. (2010). Changing gender
domination in a big four accounting firm: flexibility, performance and client
service in practice. Accounting, Organizations and Society, 35, 775-791.
Kossek, E., Barber, A., and Winters, D. (1999a). Using flexible schedules in the
managerial world: the power of peers. Human Resource Management, 38,
33-46.
Kossek, E., Noe, R., and DeMarr, B. (1999b). Work-family role synthesis:
individual and organizational determinants. International Journal of
Conflict Management, 10, 102-129.
Kossek, E., and Ozeki, C. (1999). Bridging the work-family policy and
productivity gap: a literature review. Community, Work and Family, 2, 7-
32.
KPMG (2018). Women’s leadership study: Moving women forward into
leadership roles. Retrieved from
https://home.kpmg.com/content/dam/kpmg/ph/pdf/ThoughtLeadershipPubl
ications/KPMGWomensLeadershipStudy.pdf .
Krausz, M. and Freibach, N (1983). Effects of flexible working time for employed
women upon satisfaction, strains, and absenteeism. Journal of Occupational
Psychology. 56, 155-159.
Krishnan, G.V. and Parsons, L.M. (2008), “Getting to the bottom line: an
exploration of gender and earnings quality”, Journal of Business Ethics, 78,
65-76.
Kyaw, K., Olugbode, M., and Ptracci, B. (2015), Does gender diverse board mean
less earnings management? Finance Research Letters, 14, 135-141.
Journal of Business and Accounting
103
Lee. M., MacDermid, S., Williams, M., Buck, M., and Leiba-O'Sullivan, S. (2002).
Contextual factors in the success of reduced-load work arrangements among
managers and professionals. Human Resource Management, 41, 209-223.
Lyness, K.S. and Thompson, D.E. (1997). Above the glass ceiling? A comparison
of matched samples of female and male executives. Journal of Applied
Psychology, 82, 359-375.
Lupu, I. (2012). Approved routes and alternative paths: the construction of
women’s careers in large accounting firms. Evidence from the French big
four. Critical Perspectives on Accounting, 23, 351-369.
MacKenzie, S., Podsakoff, P., and Fetter, R. (1991). Organizational citizenship
behavior and objective productivity as determinants of managerial
evaluations of salespersons’ performance. Organizational Behavior and
Human Decision Processes, 50, 123–150.
Mathieu, J. E. and Zajac, D. M. (1990). A review and meta-analysis of the
antecedents, correlates and consequences of organizational commitment,
Psychological Bulletin, 108, 171-194.
McGuire, J.B. and Liro, J.R. (1987). Absenteeism and flexible work schedules.
Public Personnel Management, 16, 47-59.
Miller, C.C., Quealy, K., and Sanger-Katz. (2018). 500 Fortune 500 Chief
Executives who are women, 500 Fortune 500 Chief Executives named John.
New York Times, April 25, 2018, 1.
Nord, W., Fox, D., Phoenix, A., and Viano. K. (2002). Real-world reactions to
work-life balance programs: lessons for effective implementation.
Organizational Dynamics, 30, 223-238.
Parker, L. and Allen. T. (2001). Work/family benefits: variable related to
employees' fairness perceptions. Journal of Vocational Behavior, 58, 453-
468.
Parker, L. (2008). Strategic management and accounting processes:
acknowledging gender. Accounting, Auditing & Accountability Journal,
21(4):611-31.
Pierce, J.L. and Newstrom, J.W. (1982). Employee responses to flexible work
schedules: An inter-organization, inter-system comparison. Journal of
Management, 8, 9-25.
Pierce, J.L. and Newstrom, J.W. (1983). The design of flexible work schedules and
employee responses: Relationships and process. Journal of Occupational
Behavior, 4, 247-262.
Powell, G., and Mainiero, L. (1999). Managerial decision making regarding
alternative work arrangements. Journal of Occupational and Organizational
Psychology, 72, 41-56.
PwC. (2018a). PwC Time to talk: What has to change for women at work.
Retrieved from https://www.pwc.com/timetotalk .
PwC. (2018b). PwC Women in Work Index, Closing the gender pay gap. March
2018. Retrieved from https://www.pwc.co.uk/economic-
services/WIWI/women-in-work-index-2018.pdf .
Anderson and Smith
104
Rhodes, D. (2001). Balanced lives: Changing the culture of legal practice.
American Bar Association: Commission on Women in the Profession,
Chicago.
Rogier, S. and Padgett, M. (2004). The impact of utilizing a flexible work schedule
on the perceived career advancement potential of women. Human Resource
Development Quarterly, 15, 89-106.
Rothausen, T (1994). Job satisfaction and the parent worker: The role of flexibility
and rewards. Journal of Vocational Behavior, 44, 317-336.
Rothausen, T., Gonzalez, J., Clarke, N., and O'Dell, L. (1998). Family—friendly
backlash—fact or fiction? The case of organizations' on-site child care
centers. Personnel Psychology, 51, 685-706.
Scandura, T. and Lankau, M. (1997). Relationships of gender, family
responsibility and flexible work hours to organizational commitment and job
satisfaction. Journal of Organizational Behavior, 18, 377-391.
Schwartz, D. (1994). An examination of the impact of family-friendly policies on
the glass ceiling. New York: Families and Work Institute.
Smithson, J. and Stokoe, E. (2005). Discourses of work-life balance: Negotiating
'genderblind' terms in organizations. Gender, Work and Organization, 12,
147-168.
Srinidhi, B., Gul, F., and Tsui, J. (2011), “Female directors and earnings quality”,
Contemporary Accounting Research, 28, 1610-1644.
Thomas, L.T. & Ganster, D.C. (1995). Impact of family-supportive work variables
on workfamily conflict and strain: A control perspective. Journal of Applied
Psychology, 80, 6-15
Trinidad, C. and Normore, A.H. (2005). Leadership and gender: a dangerous
liaison? Leadership & Organization Development Journal, 26, 574-90.
Vandenberg, R. J. and Lance, C. E. (1992). Examining the causal order of job
satisfaction and organizational commitment. Journal of Management, 18,
153-167.
Wayne, J. and Cordeiro, B. (2003). Who is a good organizational citizen? Social
perception of male and female employees who use family leave. Sex Roles,
49, 233-246.
Welsch, H. P. and LaVan, H. (1981). Inter-relationships between organizational
commitment and job characteristics, job satisfaction, professional behavior,
and organizational climate. Human Relations, 34, 1079-1089.
Whiting, R.H. and Wright, C. (2001). Explaining gender inequity in the New
Zealand accounting profession. British Accounting Review, 33, 191-222.
Williams, L. J. and Hazer, J. T. (1986). Antecedents and consequences of
satisfaction and commitment in turnover models: A reanalysis using latent
variable structural equation methods. Journal of Applied Psychology, 71,
219-231.
Wilson, F. (1995). Organizational Behavior and Gender. McGraw-Hill, London.
Journal of Business and Accounting
Volume 12, No 1; Fall 2019
105
VARIATIONS IN UNFUNDED PENSION
LIABILITIES ACROSS U.S. STATES
Paul A. Tomolonis Western New England University
ABSTRACT
Unfunded state pension liabilities are a rising concern for states and
taxpayers as the bills from previous promises are coming due, and investment
returns are at low point. The provenance of these looming obligations can help us
understand how to mitigate this issue. This paper demonstrates a lack of
association of a state’s ability to pay as a driver of unfunded pension liabilities
(gross state product or tax revenues). Further, this work provides evidence that
service levels (state employees per capita) and promises made to state retirees
(total liability per member) are potential drivers of unfunded public pension
liabilities. Given these results, we might conclude that state governments are their
own worst enemy in the generation of these unfunded pension liabilities, especially
since there is evidence suggesting that these deals were crafted for political
expedience such as delaying the inevitable facing of fiscal imbalance (debt and
deficits correlated with unfunded pension liabilities). Further complicating this
issue is the possible flight of the tax base of a given state as the residents
(taxpayers) discover the truth about the looming fiscal issues and conclude that a
more prudent state government would better serve them.
Key words: Pensions, unfunded pension liabilities, state pensions, pension
obligation
INTRODUCTION
In 2013, US states reported a total unfunded pension costs for state
employees of $968 billion or 6.9% of aggregate state income (Pew 2017). Pension
liabilities arise from state promises to its current and former employees (state
workers) to fund their retirement, often including healthcare coverage or other
postemployment benefits (OPEB). Unfunded pension liabilities represent the
estimated portion of this future cost (liability) that is not set aside (invested) today
based on estimated future payouts and investment returns over the relevant time-
period. To be clear, the $968B in unfunded pension obligations cited here excludes
OPEB. If included, it is estimated to add an additional 28% to this amount at the
state level (Munell, Aubry & Crawford 2016). To estimate unfunded pension
liabilities, actuaries consider several uncertain aspects of a given pension plan’s
members; such as their current ages and expected lifespans as well as those of their
Tomolonis
106
spouses, because pension payments often continue until death of both household
members. These pension characteristics can vary by plan (across and within states)
and are are accounted for by actuaries that deal with statistical predictions from
demographic trends.
The generosity of survivor benefits (percent of payment that continues
after the death of the primary beneficiary) is where the promises of politicians first
start to affect actuarial distributions of these defined benefit programs. A common
theme among the politicians that seek election (and re-election) from their
constituents is to make the present seem better or improved at a low cost. This can
be achieved by a sleight of hand regarding the future since people tend to discount
future costs more heavily than future benefits, especially when the cost is borne by
society while the benefit accrues to individuals. Further, by shifting costs to future
periods, balancing the state budget (required in several states) is made easier.
Survivor benefits are just the tip of the iceberg in state pensions since
current negotiations with state employee unions include retirement age, years of
service calculations, average wage calculations (e.g. weighting more recent pay
higher), cost of living adjustments, and various other “credits” that improve the
future payout of pensions, often in return for lower wages today. These future
benefit promises are easier to ignore since their cost can be shifted to future periods
and further obfuscated by over-estimation of pension plan return on investment.
Of course, states must invest in order to expect any return, and this is where
unfunded pension liabilities create the trade-off between future benefits and
current costs. State governments often make optimistic assumptions about
investment returns and actuarial estimations compensate with conservative
assumptions in their predictions about the future costs of administering state
pension plans. Actuarial reports regarding the “health” of a given pension plan are
commissioned by the states on a regular basis, although not required or consistently
formed (estimated). A “healthy” plan is often regarded as one that is at least 90%
funded. However, the true health of a pension plan can be difficult to judge
accurately because of a combination of a lack of consistent standards of
measurement and a lack of actuarial independence, both under the auspice of the
state government.
The funding status of a state’s pension fund is considered fully funded
when, based on actuarially estimated returns and payouts, the fund has enough
current investments to pay the future anticipated costs (note that current
contributions to the fund should be invested and not used to pay current retirees
who are paid from the funds returns and principal). Ideally, any pension fund
would set aside the anticipated costs when the employee performs the work and
thus earns the right to the future cash flow (retirement benefit) but, as with all
budgets, current costs appear more significant than distant future payments. Add
to this time-horizon issue the overly optimistic estimates of low payouts along with
equal or greater optimism for high investment returns and the results are low
Journal of Business and Accounting
107
perceived current costs (pension investments set aside) with rising future
obligations that are unfunded, even in the actuarial (conservative) sense. While
this appears to a significant issue for the states, a state by state comparison reveals
that the health of state pension plans varies dramatically from fully funded
(actuarially) to substantial looming costs (unfunded) to keep up with future (and
current) expectations of pension members (Barkley 2012). So, what causes this
disparity across the states?
Funding Disparity
As with any budget, an imbalance can be analyzed from the flow of
resources available to finance spending, from the flow of resources spent
(promised), or both. Since the financial crisis of 2008, states have recovered at
different rates impacting their ability to collect tax revenue (Pew 2017). Tax
revenue and the source of that revenue (state income or gross state product) are
simple enough to measure with Census data. On the spending side, promises made
to fund pension plans can arise from the level of state services provided or from
the promises made to the state employees (or both). These spending side levels can
be approximated with the number of state employees per capita (as a proxy for the
level of state services) and total pension liability per member of the pension plan
or per capita. Since both low inflows and high outflows could be happening
simultaneously, it is natural to begin with the association between state resources
and total pension liability.
While correlation does not imply causality, if total pension liability is
positively correlated with state resources (ability to pay) then perhaps the “rich”
states are over-promising to their employees because they feel rich. If total pension
liability is negatively correlated with a state’s ability to pay then perhaps the state
is making pension promises in lieu of an ability to compensate state employees in
the current period and that might lead to greater underfunding of the pension
obligation (liability). If there is no correlation, then the question of underfunding
may be undetected or unrelated to ability to pay and our focus should turn to
promises made. These posited relationships are summarized in Chart 1.
CHART 1: Correlations and possible causes
Positive Correlation: Ability to Pay
(GSP) & Total Pension Liability
Rich states remunerating state
employees from state largesse
Negative Correlation: Ability to
Pay (GSP) & Total Pension Liability
Poor states using pension promises to
forestall current budget issues
No Correlation: Ability to Pay (GSP)
& Total Pension Liability
Ability to pay not driving pension
funding status, check promises made
Tomolonis
108
Using 2013 data adjusted for cost of living differences among the states,
there appears to be no correlation (or ever so slight positive correlation) between a
state’s ability to pay and its total pension liabilities (as discussed next).
Figure 1 shows a scatter plot of a state’s total pension liability per capita
versus gross state product (income) per capita by state.
FIGURE 1: Total Pension Liability per capita v GSP per capita with median
The graph (Figure 1) shows all states and has a median GSP of about
$44,000 and total pension liability per capita of about $10,000 (bold vertical and
horizontal lines) dividing states into high and low ability to pay (above or below
median GSP) and high and low promises of state services (above or below median
pension liability per capita). Adding a trend line to this graph results in a R2 of
about 7.45% or very little correlation between a state’s ability to pay and it pension
liability per capita.
Removing the outliers of Alaska, Hawaii, Vermont, Delaware, and Wyoming,
reduces the R2 to less than 0.1% (0.06%) and allows a clearer focus on the
Alaska
California
Delaware
Hawaii
Illinois
Massachusetts
Mississippi
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Dakota
Ohio
Oregon
Texas
Vermont
Wisconsin
Wyoming
R² = 0.0745
$5,000.00
$7,000.00
$9,000.00
$11,000.00
$13,000.00
$15,000.00
$17,000.00
$19,000.00
$21,000.00
$23,000.00 $33,000.00 $43,000.00 $53,000.00 $63,000.00
GSP/cap
Liab/Cap
Journal of Business and Accounting
109
remaining states. Again, there appears to be no correlation between pension
liabilities and a state’s ability to pay for them in a per capita sense. There does
appear to be larger UNFUNDED liabilities as the total liability per capita grows,
but this should come as no surprise since there are fewer taxpayers to foot the bill.
Investigating the relationship between total liabilities per pension plan
member and gross state product has similar results (i.e. little correlation). Here the
interpretation on ability to pay is the same, but now considering the total pension
liability per member of the plan is used as a proxy for the level of promises made
to the plan members for future benefits. The full panel of states has an R2 of 1.4%,
with a median total pension liability per member of over $127,000. Removing the
outliers of Nevada, Delaware, Hawaii, and Alaska, raises the median promise level
to $130,000 and slightly increases the R2 to 1.7%, confirming what can be seen in
both approximations: no correlation between a state’s ability to pay and its
promises per member of the state work force. Interestingly, some of the largest
unfunded pension liabilities are in the high ability to pay, high promises made to
state workers quadrant when separating by median ability (GSP per capita) and
median promise (pension liability per capita).
These preliminary results leave open the question of how state resources
(ability to pay) affect the funding status of the state pensions. It could be that states
with lower ability to pay cannot afford to meet similar obligations for pension plans
as other states with more resources leading to hypothesis 1:
Hypothesis 1: States with a lower ability to pay will face higher unfunded
pension liabilities.
It is also possible that state governments make promises of higher levels
of state service for the broader electorate to curry favor at re-election time or to
ameliorate some real or perceived social inequity.
Hypothesis 2: States with a higher percentage of state employees will
face higher unfunded pension liabilities.
It might be the case that state governments make more significant pension
promises to state employees as a form of deferred compensation or to appease a
particularly vocal constituency (state employees).
Hypothesis 3: States with higher level of total pension liabilities will face
higher unfunded pension liabilities.
This last hypothesis has an obvious mechanical nature to it since higher
levels of pensions promises are likely to lead higher unfunded pension liabilities,
but controlling for ability to pay and taking a longer-term perspective allows for
some additional insight.
Tomolonis
110
As a corollary to Hypothesis 3, states might provide additional pension
(deferred) compensation in exchange for lower current period costs if the state is
in a fiscal bind, especially if the state budget is required to be balanced by state
law.
Hypothesis 4: States with more significant budget issues (deficits and
debt) will face higher unfunded pension liabilities.
As preliminary results, the level of state resources (income or tax revenue)
reveals no correlation between the source of funding and unfunded pension
liabilities despite the wide variance in unfunded liabilities across states. The results
from level of state services (state employees per capita) shows weak correlation
with unfunded pension liabilities; whereas data on promises made to state
employees (total pension liability per plan member) reveals stronger correlations
despite the confounding issues of the time members are in the state pension plan
and the total level of membership. Results on the level of debt per capita and
deficits both indicate a strong correlation with unfunded pension liabilities, which
should come as no surprise since robbing the future is a common way to pay for
the present.
After a brief overview of the literature in this area, this paper will discuss
the methodology used to address the potential causes of variation in funding status
across the states, offer a summary of the results, and provide some concluding
remarks regarding the causes and possible avenues to address these issues. The
hope is that states that are in funding trouble might be able to learn from the states
that are funded, unless there are significant differences that preclude the
application of these results.
LITERATURE REVIEW
The Pew reports (2017, 2014) are good at quantifying and comparing the
state pension funding issues, but are only a part of the problem. Lutz & Sheiner
(2014) estimate that unfunded OPEB might be as much as an additional half the
size of unfunded pensions (despite being about a quarter of the overall cost) and
demonstrate a similar variation across the states. OPEB unfunded obligations are
often easier to “hide” since there are more estimations and therefore more opacity
to the situation as a whole. Just on the pension side, even the actuarial estimates
forecast an increase in ARC (annual required contribution) to keep the pension
system current (Brainard & Brown 2015). These actuaries are paid by the state
running the fund and thus, even their favorable estimations are based on statistical
analysis of past trends that do not fully incorporate current financial market
expectations as used by the financial professionals that manage the pension funds’
investments. In fact, Rauh & Novy-Marx (2010) estimate that even with zero cost
of living adjustment and a move to the Social Security retirement age (many
pensions allow for early retirement after 20 or 30 years of service) using Treasury
Journal of Business and Accounting
111
Bill discount rates would put the pension deficit at $1.5 Trillion rather than the
actuarially estimated $968 Billion. Liu, Lu & Zheng (2010) perform a similar
ranking of pension status with qualitatively similar results to Pew and while
actuarial discounting might understate the problem, it does not appear to change
the relative ranking of pension funds.
While we can all agree that a problem exists and that it clearly varies by
state, there is more of a paucity of research on the choices as well as the causes.
Gale & Krupkin (2016) argue that options are limited since these pension (and
collective bargaining) agreements are binding as well as being a political hot
button. Abrogation of the contracts through bankruptcy is the threat point to
engender renegotiation of the contracts but the lack of political will to upset the
system that the politicians rely on to deliver services to their constituents as well
as the state employee constituency itself does not favor reform. Yet reform such as
concessions, increases in taxes, or cuts in state service levels is necessary to meet
the obligations in a way that is fair to all. Even as far back as Porterba (1996) we
see evidence of short term budgeting at the state fiscal level that favors current
constituents by passing the costs on to future generations. Porterba noted that many
states have a balanced budget amendment in their state constitutions and a popular
way to get around that restriction is through pension promises for tomorrow to get
lower wage increases today. Compounding these issues are the downward biased
actuarial estimates of the unfunded pension obligation generated by using
upwardly biased estimations of long-run returns to the pension assets. Johnson
(1997) finds that states are using these mis-estimations to fund current period lower
taxes and allow for generous pension promises that are unfunded today. Cheney,
et al (2003) provide further empirical evidence of states underfunding and or mis-
estimating due to budget constraints in taxes or spending. Those mis-estimations
in actuarial assumptions are associated with states that have most fiscal difficulty
according to Eaton and Nofsinger (2004) and this extends into private pensions (as
would be expected) according to Comprix and Muller (2006).
Missing from this stream of literature is additional research on the
fundamental causes of the pension problem. Is it just political expedience that
creates this issue, or are more basic economics such as the ability to pay or the
level of services at the root of these funding problems? Exposing these issues more
clearly provides insight so that we can learn from past mistakes and not fall prey
to the same issues as we try to repair the current system. Isolating the economic
cause allows those that will face this issue (future generations of citizens) to make
wiser choices in the politicians they elect and the places they choose to live.
METHODOLOGY
With so many states facing unfunded pension obligations, it raises the
question: whence do these issues arise? We can think of it in two ways: either the
state wishes to provide services for its residence on par with other states but cannot
Tomolonis
112
afford it (ability to pay) or states that have the ability to pay are making promises
of higher pensions to their state employees. The latter could be either a case of
pension promises that are consistent with a higher level of state services to
residents (e.g. more state employees per capita). A more nefarious motivation such
as to defer current pay (or pay increases) in order to balance a fiscal budget today
or even a political side payment for election support, is a possible future area of
research. In order to attempt separating and testing these possibilities we can think
of states as either high or low ability to pay states (HA, LA) as well as either high
or low promise of future pay states (HP, LP). Since these conditions are not
mutually exclusive we have a two by two categorization of states, represented in
Chart 2 below.
CHART 2: State public pension characteristics (ability to pay and
promises made to workers)
High Ability (HA), High Promise (HP) High Ability (HA), Low Promise (LP)
Low Ability (LA), High Promise (HP) Low Ability (LA), Low Promise (LP)
To discern a state’s ability to pay both Gross State Product (GSP) and Tax
Revenues (TR) are considered (Census Bureau data). Gross State Product is, of
course, a measure of the total income of a state and should be highly correlated
with ability to pay for state worker pensions. The confounding factors include
states with porous borders (small states or states with significant portions of the
population at a border) especially those with high tax rates relative to the bordering
states. To account for this, tax revenues collected by the states is tested for
association with unfunded pensions and this results in qualitatively similar results
in all cases. Tax revenues are clearly dependent on the tax rate(s) in each state, but
tax arbitrage (competition) among the states (especially those with porous borders)
limits the amount of variance to some degree; so, the similar results should not be
surprising.
The ability to pay results indicate that the level of unfunded pension
obligations (Pew Report) in a state is, at best, unrelated to both measures of ability
to pay (when adjusting for the cost of living between states). Further, unfunded
pension liabilities are actually positively correlated with ability to pay when using
unadjusted (CoL) levels (GSP: 0.1 p=0.011, TR: 0.6, p=0.031) which is counter to
the expectation (Pew Report). Thus, it can be concluded that relatively poor states
with less ability to pay for state services provided by state employees are not a
significant driver of unfunded pension liabilities.
So then, what is driving some states further into an unfunded status while
others are able to fully (actuarially speaking) fund future pensions obligations. One
possible explanation is that states make “promises” to state employees for either a
higher level of state services (i.e. lower pay today, but more generous pensions) or
for political expedience such as delaying compensation in order to balance the
Journal of Business and Accounting
113
current fiscal period’s budget. Regardless of the reason, let us consider how to
measure these promises made and then attempt to disentangle the potential causes.
In this study, the level of promises made to state employees is proxied by
the total pension liability per capita (LpC) and total pension liability per member
of the pension plan (LpM), which includes past and present employees. Total
pension liability means that this is the total pension cost for future payments based
on past employment (as actuarially determined), not just the unfunded portion of
that liability. Higher pension liability per capita suggest that state employees are
paid and or promised more future pay, which could result from higher state
services being provided or a higher cost of living. To compensate for the cost of
living differences the Census Bureau regional price deflators by state are used to
adjust the data. Higher total liability per capita is also consistent with higher
deferred compensation in exchange for pay concessions today (reduce current
fiscal pressures) which is more likely to correlate with higher unfunded portions
of this liability. The data show that total liabilities per capita (LpC) has significant
(p=0.000) correlation with unfunded pension liabilities (point estimates from 0.41
to 0.35) when using both cost of living adjusted data and unadjusted data and
whether or not including GSP as a control for ability to pay. This suggests that the
level of promises made to state employees is a driver of unfunded pension
liabilities regardless of the relative cost of living or state’s ability to pay (overall
or per capita).
It could also be the case that there are more state pension members
(consistent with a higher service level) so total pension liability per plan member
is also considered. Pension liability per plan member rises as the level of promises
to each member increases, which is a more direct measurement of the promises,
but with less connection to state services levels provided to the state’s citizens.
However, this measure also suffers from diffusion over a longer time frame: it
includes as members those just hired all the way out to those who might have been
retired for 20 or more years. The level of promises is likely to have varied greatly
over that time frame and more recent hires are in the divisor (denominator) despite
not contributing much to the dividend (numerator). Despite this bias, the LpM still
produces weakly significant point estimates (within the 10% range: p=0.003 for
uncontrolled, unadjusted data to p=0.076 for GSP controlled and cost of living
adjusted data with point estimates from 0.01 to 0.02) indicating correlation to
unfunded pension liabilities. This suggests that these unfunded liabilities are
partially correlated with promises made to state employees.
To further explore the relation between these promises made to state
employees and their causes, the level of services provided is estimated by the
number of current state employees per capita and the total state retirement plan
membership per current capita. The number of current (EE/capita) or current and
former (Member/capita) state employees per capita is suggestive of the level of
services provided but could be confounded with low productivity of employees (to
the extent that varies by state) or by excessive hiring as a concession to state unions
Tomolonis
114
for lower current fiscal period pay increases to existing members. Of course, we
would expect higher levels of members per capita to be associated with higher total
pension cost and higher unfunded pension liabilities as these higher costs are
spread among fewer residents, and the results are significant (p=0.000,
point=77,713). However, to the extent that prior state employment levels are
correlated with current state employees (consistent with the level of service
argument for higher pension costs), we would expect higher current employment
levels to also be associated with higher pension costs and greater unfunded pension
liabilities. This is a reasonable assumption barring any significance difference in
longevity or retirement age (an open question relating back to promises made for
political expedience). The results indicate a break between current and former plan
members since the liability per capita is not correlated with current state employees
per capita (p=0.27). This break suggests that the level of service argument is not
supported as a reason for a state to have excessive unfunded pension liabilities.
RESULTS
Estimating the link between ability to pay and unfunded liabilities suggests
a simple correlation model:
(1) Unfunded level = α + β(ability to pay) + ε
With the unfunded level and ability to pay measures normalized by
population (per capita); additionally, cost of living adjustments by state are made
using Census deflators.
RESULTS 1: Unfunded pension liabilities on GSP or Tax Revenue (ability
to pay)
Dep.Var. =
unfunded
pension
liabilities per
capita
Intercept β CoL adjusted β
GSP per capita (p
value)
-1162.57 (0.50)
1924.43 (0.257)
0.098 (0.012)**
0.026 (0.494)
Tax Revenue per
capita (p value)
1720.84
(0.032)**
2379.08
(0.002)***
0.0005
(0.041)**
0.0003 (0.306)
To compare with prior research on political and fiscal motivations for
unfunded pension liabilities (Chaney et. al, 2003; Eaton & Nofsinger 2004) the
fiscal stress of a state is considered as a driver of unfunded pension liabilities.
Journal of Business and Accounting
115
When the level of unfunded pension liability per capita is compared with the level
of state debt or current year deficits per capita (and controlling for income as well
as using cost-adjusted data) higher debt or deficits are significant drivers of
unfunded liabilities (p=0.026 and p=0.049, respectively).
Thus, no association between ability to pay (state resources) and unfunded
pension liabilities appears, similar to the earlier results regarding total pension
liabilities and a state’s ability to pay measures; Hypothesis 1 fails. That re-opens
the question as to what does cause some state’s to run into pension funding trouble
while others seem to be able to fully fund their state pensions.
Estimating the link between the level of state services to the public and
unfunded liabilities is accomplished with state employees per capita and state
pensions members per capita (past level of services) while controlling for ability
to pay:
(2) Unfunded level =α + β(level of service to public) + γ(ability to pay
control) +ε
RESULTS 2: Unfunded pension liabilities on number of state workers or
pension plan members
Dep.Var. =
unfunded
pension
liabilities per
capita
Intercept β ᵞ
State Emp per
capita (p value) -2,480 (0.1498) 119,939 (0.0128) 0.08 (0.0280)
Plan members per
capita (p value) -2,888 (0.1319) 22,500 (0.0578) 0.09 (0.0133)
State Emp per
capita - CoL
adjusted (p
value)
115 (0.95417) 82,368 (0.0641) 0.04 (0.3480)
Plan members
per capita-CoL
adjusted(p value)
425 (0.8024) 27,170 (0.0127) 0.01 (0.8319)
It appears that there is at least a weak association between the level of state
services provided and the unfunded pension liabilities. So, appeasing state voters
by providing a higher level of services and then delaying the payment for these
services is at least mildly suggested. The level of state services as measured by
total pension plan members (retired and active) per capita is more strongly
Tomolonis
116
associated (at the 5% level) with unfunded pension liabilities, suggesting that at
least part of the unfunded liability is a result of the state government’s provision
of services. This provides some support for Hypothesis 2 but using total plan
members per capita as a measure of state services suffers from the possibility of
changes in the denominator (population) driving the results. For example, as states
make increasing high promises to state workers in the future (pension promises) it
drives up the expected cost and hence likely future burden on taxpayers. To the
extent that these taxpayers are mobile, they might move to states with a lower
expected future burden on taxpayers (e.g. migration from the Northeast to the
Southeast that has been witnessed over recent decades).
So then, if a state’s ability to pay is unrelated to unfunded pension
liabilities and the level of services provided is only weakly associated with them,
it might be that states are promising their employees too much in the future,
possibly to defer that cost. Using the total level of pension liabilities (funded and
unfunded) as a measure of these promises (both per capita and per plan member)
provides some insight into the relationship between pension promises and
unfunded pension liabilities:
(3) Unfunded level = α + β(pension promises) + γ (ability to pay control)
+ ε
RESULTS 3: Unfunded pension liabilities on total pension liabilities
Dep.Var. =
unfunded pension
liabilities per
capita
Intercept β Γ
Total liability
per capita (p
value)
-1,986 (0.1644) 0.3867 (0.0000) 0.021 (0.5267)
Total liability per
plan member (p
value)
-1,290 (0.4395) 0.015 (0.0453) 0.054 (0.2020)
Total liability
per capita - CoL
adjusted (p
value)
132 (0.9306) 0.3611 (0.0000) -0.014 (0.6814)
Total liability
per plan member
-CoL adjusted (p
value)
260 (0.8902) 0.015 (0.0764) 0.019 (0.6141)
The results of the pension promises show strong correlation on a per capita
basis, but the natural bias there is towards correlation since higher total liability
Journal of Business and Accounting
117
per capita is more likely to drive higher unfunded liability per capita, even when
controlling for a state’s ability to pay. The weaker statistical significance (at the
10% level) with total pension liability per plan member also biased, but away from
results since the higher promises are likely to be the more current agreements
whereas the older agreements (with current retirees) are the ones that would drive
higher unfunded pension liabilities given the longer time horizon of accumulation
and the shorter time horizon for investment returns. When taken together, these
results suggest that the level of promises made to state workers is weakly
associated with higher unfunded pension liabilities under Hypothesis 3.
The question of why state governments would make these pension
promises to state employees is still unanswered. One possibility is the deferral of
compensation for state employees to make the state budget easier to balance
(especially if it is required to balance by law). Some light may be shed on this by
considering the association between state debt (or deficit) levels and unfunded
pension liabilities.
Estimating the connection between budget shortfalls and unfunded
pension liabilities can be done with state debt per capita and state deficits
controlling for a state’s ability to pay since debt might be substituted for a state’s
ability to pay:
(4) Unfunded level = α + β(debt or deficit) + γ (ability to pay control) + ε
RESULTS 4: Unfunded pension liabilities on state debt or deficit
Dep.Var. =
unfunded
pension
liabilities per
capita
Intercept β Γ
State debt per
capita (p value) 162 (0.9252)
0.365
(0.0184) 0.037 (0.391)
State deficit (p
value) -2,887,624 (0.1318) 22,500 (0.0578) 94.07 (0.0133)
State debt per
capita
CoL adjusted
(p value)
488 (0.7782) 0.370 (0.0260) 0.0298 (0.4199)
State deficit
CoL adjusted (p
value)
1,434,401 (0.3871) 18,970 (0.0492) 1.13 (0.9769)
From these results there appears to be strong support for the association
between state debt (or deficit) and unfunded pension liabilities as hypothesized
Tomolonis
118
(Hypothesis 4). It should be noted that while debt per capita produces significant
results, deficit per capita was not even close (p=0.756), yet deficit level versus
unfunded pension liability per capita was significant (see tabulated results). This
could be the result of high deficit states looking to balance the budget with an
extraordinary move like slowing funding payments to pensions whereas low deficit
states turn to other measures. That result is obfuscated by dispersing the current
deficit over more citizens. Also, the recency of deficits versus debt could be a
factor as well: a deficit this year might not change the contribution made to the
pension plan very much whereas a series of deficits resulting in higher debt levels
(per capita) likely has a higher cumulative effect on unfunded pension liabilities
and states with a higher current deficit (in total) are more likely to have higher debt
levels (long term deficits). In other words, there are high and low spending states
but some of the high spending states have a higher population and so get shifted
into the middle on a per capita basis and these high spending states are the one
having trouble funding their pensions regardless of population.
Given the results (summarized in Chart 3) from the tests of association
between unfunded pension liabilities by state and the possible causes, it seems that
funding is within the control of the state governments.
CHART 3 RESULTS SUMMARIZED by hypothesis
Unfunded
pension
liabilities are
associated
with:
Measure: Significance: Result:
H1: a state’s
ability to pay
Income (GSP) pC
Tax revenue pC
None
None Fail
H2: a state’s
level of
service
provided
State employees
per capita (pC)
Plan members pC
10% level
5% level
Weak
pass
H3: a state’s
promises to
its pension
plan
Total pension
liability per capita
Total liability per
member
1% level
10% level
Pass
H4: a
state’s budget
balance
State debt per
capita
State deficit
5% level
5% level
Strong pass
Journal of Business and Accounting
119
So what does the future hold for these various pension plans with such
wide differences in funding status across states?
Obviously, states that are fully funded must keep up with the future retirees
but states that are behind in funding today must catch up or face the possible
migration of their tax base away from the future funding. These states might catch
a break if the number of state employees retired slows or reverses (this would also
make funded states have an even brighter future); however, if the upcoming
number of retirees is rising then states that are behind in funding will face even
more difficulty in funding, exacerbating the demographic migration threat. You
can only raise state taxes so much before your tax base moves to another state (tax
arbitrage) given the relative ease of moving. In fact, once this trend starts, it builds
on itself as the best and brightest seek opportunities and congregate in new areas,
bringing tax dollars, productivity, and output with them. This leaves the unfunded
and worsening states in a terrible predicament that might necessitate the re-
negotiation of state pension agreements or concessions from state workers unions.
Even those fully funded states might be subject to these same market (the market
for state services are a reasonable cost) forces as their future pension liabilities rise.
At least those states have choices in their future, but the states with the issue today
must solve the Laffer curve trade-off between tax rates and growth quickly or risk
losing their tax-paying base.
To visualize this issue, consider states along two dimensions: the funding
status of the current pension liabilities and the present level of state retirees per
capita. This suggests a two by two matrix of possibilities (Chart 4) in to which we
can map the various states by their current status.
CHART 4: State pension status by median (funding, current retirees)
Low Funding, Low Retirees Low Funding, High Retirees
High Funding, Low Retirees High Funding, High Retirees
Subtracting the current state employees per capita form the total state
pension membership per capita, an estimate of the number of retired state workers
per capita is found. Observing (see Figure 2) a clustering of states by unfunded
liabilities and the estimate of state pensioners per capita we can see four groupings
(quadrants), with low unfunded status and many retirees (lower right, e.g.
Wisconsin) as being the least risky group since they are paid up for those that are
retired now (little uncertainty about the bill). Next, funded states that have few
current retires (lower left, e.g. Tennessee) are current and have the flexibility to
stay out of trouble by not over-promising today. The upper right quadrant (e.g.
Mississippi) have a high number of current retirees and a low funding status, so
they are in current trouble but working through it; if they can pay now and avoid
Tomolonis
120
FIGURE 2: Unfunded Liabilities per capita v. retired state employees per
capita
Alaska
Arizona
California
Colorado
Connecticut
Delaware
Florida
Hawaii
Illinois
Iowa
Louisiana
Maine
Massachusetts
Minnesota
Mississippi
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Dakota
Ohio
South Carolina
Tennessee
Texas
Vermont
Washington
Wisconsin
Wyoming
-800
200
1200
2200
3200
4200
5200
6200
7200
8200
9200
0.03 0.05 0.07 0.09 0.11 0.13
Unfunded Liability per capita
State Retireesper capita
Journal of Business and Accounting
121
making promises for the future they can head in the correct direction. The upper
left quadrant (e.g. Connecticut) has a low funded status and few retirees, so their
problems lie ahead, especially if they continue to make promises they cannot keep.
The funding status of state pensions is clearly a choice of the state
government, but so is the number of current state employees that will become
future retirees. Assuming a normal trend across states with reversion to the mean,
those states with a high level of current retirees are likely to see lower levels in the
future, and those with lower levels are likely to see a rise, unless the state
government of the given state is entrenched in its ways which might cause the trend
to continue of worsen. Factors such as these are what lead to the tax arbitrage
migration discussed in the preceding analysis.
So what drives states to make these promises? Political expedience in
balancing the state’s budget has been proffered as a reason. The test of this
relationship suggests that states that are in fiscal trouble have higher unfunded
liabilities. This is not surprising since if the state cannot balance its budget it is
likely to have difficulty in setting aside money for long-term obligations. Yet, the
compounding of the fiscal problems by under-funding the promises made to state
employees is, in part, masking the fiscal problems of the state, especially if the
worst days (most retirees) lie ahead.
CONCLUSION
When taken together, the lack of evidence on a state’s ability to pay as a
driver of unfunded pension liabilities, along with the evidence that service levels
and promises made to state retirees are a potential driver, we might conclude that
state governments are their own worst enemy in the generation of these unfunded
pension liabilities. This is especially true in light of the evidence suggesting that
these deals were crafted for political expedience such as delaying the inevitable
facing of fiscal imbalance. Further complicating this issue is the possible flight of
the tax base of a given state as the residents (taxpayers) discover the truth about
the looming fiscal issues and conclude that they would be better served by a more
prudent state government. This sleight of fiscal hand is leaving those taxed by the
state and its future pensioners potentially holding the bag; unfortunately, that bag
might not be as full as promised or envisioned.
REFERENCES
Barkley, Racheal Morningstar The State of State Pension Plans – A Deep Dive
Into Shortfalls and Surpluses, 2012 Web.
Brainard, Keith, and Alex Brown National Association of State Retirement,
Administrators March. The Annual Required Contribution Experience of
State Retirement Plans, FY 01 to FY 13. 2015 Web.
Tomolonis
122
Calabrese, Thad. Public Pensions, Public Budgets, and the Risks of Pension
Obligation Bonds., 2010. Web.
Chaney, Barbara A., Paul A. Copley, and Mary S. Stone. "The Effect of Fiscal
Stress and Balanced Budget Requirements on the Funding and Measurement
of State Pension Obligations." Journal of Accounting and Public Policy 21.4
(2003): 287-313. Web.
Comprix, Joseph, and Karl A. Muller. "Asymmetric Treatment of Reported
Pension Expense and Income Amounts in CEO Cash Compensation
Calculations." Journal of Accounting and Economics 42.3 (2006): 385-416.
Web.
Eaton, Tim V., and John R. Nofsinger. "The Effect of Financial Constraints and
Political Pressure on the Management of Public Pension Plans." Journal of
Accounting and Public Policy 23.3 (2004): 161-89. Web.
Gale, William G., and Aaron Krupkin. Financing State and Local Pension
Obligations: Issues and Options. The Brookings Institution, 2016. Web.
Johnson, Richard W. "Pension Underfunding and Liberal Retirement Benefits
among State and Local Government Workers." National Tax Journal (1997):
113-42. Web.
Liu, Jiapeng, Rui Lu, and Zhengyang Zhang. "Rating the State Government
Public Pension Plans in the US." Accounting and Finance Research 4.4
(2015)
Lutz, Byron and Louise Sheiner The Fiscal Stress Arising from State and Local
Retiree Health Obligations. Journal of Health Economics 2014 (38) p. 130-
146
Marks, Barry R., KK Raman, and Earl R. Wilson. "Toward Understanding the
Determinants of Pension Underfunding in the Public Sector." Journal of
Accounting and Public Policy 7.3 (1988): 157-83. Web.
Mitchell, Olivia S., and Ping Lung Hsin. Public Sector Pension Governance and
Performance. National Bureau of Economic Research, 1994. Web.
Munnell, Alicia, Jean-Pierre Aubry and Caroline Crawford “How Big A Burden
Are State and Local OPEB Benefits” Center for Retirement Research at
Boston College, March 2016.
Journal of Business and Accounting
123
Pew Charitable Trust Report: "Fiscal 50: State Trends and Analysis." 2017 Web
Accessed 3/17 http://www.pewtrusts.org/en/research-and-
analysis/collections/2014/05/19/fiscal-50-state-trends-and-analysis
Pew Charitable Trust Report: “The State Pension Funding Gap:2014.” 2014 Web
Accessed 3/17 http://www.pewtrusts.org/en/research-and-analysis/issue-
briefs/2016/08/the-state-pension-funding-gap-2014
Poterba, James M. "Budget Institutions and Fiscal Policy in the U.S. States." The
American Economic Review 86.2 (1996): 395-400. Web.
Rauh, Joshua, Robert Novy-Marx “Policy Options For State Pension Systems
And Their Impact On Plan Liabilities” NBER Working Paper 16453
(October 2010).
Rauh, Joshua D. "Are State Public Pensions Sustainable? Why the Federal
Government should Worry about State Pension Liabilities." (2010)Web
Journal of Business and Accounting
Volume 12, No 1; Fall 2019
124
CONSISTENCY OF ETHICAL ANALYSIS
AMONG ACCOUNTING STUDENTS
Lisa Flynn
Charlene Foley Deno
Howard Buchan
SUNY Oneonta
ABSTRACT
This paper reports student responses to a series of fictitious ethical
dilemmas. We found that there was no clear indication that students use consistent
ethical frameworks when making decisions with moral content in the administered
survey situations. Additionally, we found little to no correlation between what a
student judged as important in the scenarios and the corresponding decisions that
were made. An important outcome of accounting ethics education is a student who
is equipped with the ability to make reasoned, consistent ethical decisions that
translate ethical ideals into action in real world situations, ensuring a logical and
analytical approach. Based upon our findings, that outcome has not been
completely or consistently attained. We concur that a focus on moral development
may be inadequate (see, for example, Jeffries & Lu, 2018 and Shawver & Sennetti,
2009) and suggest that ethics instruction incorporate more fully the complexities
involved in ethical decision-making. Improvements in moral reasoning may be
seen with improved classroom methods (Jones, 2009) and we encourage ethical
analysis to be viewed as a multi-faceted process that matures as the student
matures, with a corresponding focus on analytical ability that continues to develop
as situations become more complex and increasingly challenging.
Key Words: cognitive moral decision-making, ethical awareness, Defining
Issues Test II (DIT- 2), ethics instruction
INTRODUCTION
In an effort to determine the ethical thought processes and moral decision-
making abilities of our students, we administered the Defining Issues Test II (DIT-
2), which is widely used as a measure of cognitive moral decision-making, moral
reasoning, and cognitive moral development. The ethical framework used in moral
reasoning and decision-making should yield some consistent and correlated
responses between the moral statements deemed to be important in a given
scenario and the resulting choice of action. Ethical reasoning ability is a key
component in all our daily lives and is of particular importance for accounting
students that will soon be accounting professionals. As accounting professionals,
they will be charged with ensuring accurate and trustworthy financial reporting
activities. When asked to come to a decision about the appropriate action for a
Journal of Business and Accounting
125
given scenario, our findings show that students are not demonstrating an ability to
consistently apply such reasoning to diverse, ethically charged situations. The
amount and the frequency of exposure as well as the methods of delivery need
careful scrutiny to determine future approaches to ethics instruction.
BACKGROUND AND INSTRUMENTATION
Ethical analysis is often studied using the DIT-2 originally developed by
Rest (1979) as the Defining Issues Test and subsequently updated. It is grounded
in cognitive moral development theory (Kohlberg, 1969) and its three levels of
moral reasoning. The first level is labeled as pre-conventional thinking and is
characterized by punishment avoidance. The second level is labeled as
conventional thinking and focuses on following rules and laws to maintain order
for the proper functioning of society. The third level is labeled post- conventional
thinking and is characterized by principled thinking that transcends specific
societal rules in favor of universally applicable ideals. Rest (1986) expanded
cognitive moral development theory into a cognitive moral decision-making model
which posits four theoretically (though not necessarily in practice) sequential
components: moral issue identification, moral reasoning, moral intent, and moral
action.
The DIT-2 measures Rest’s (1986) second component, moral reasoning,
and characterizes results according to level of moral reasoning (pre-conventional,
conventional, post- conventional) as described by Kohlberg (1969). It provides five
ethical scenarios and lists several statements of ethical reasoning or justification
related to the given scenario. Respondents rank the list of statements according to
what is most important in choosing an action, and then respondents choose the
appropriate action for the scenario. The rankings provide data to determine the
level of moral reasoning, and the action chosen also indicates a respondent’s level
of moral reasoning.
Studies showing links between moral development scores and ethical
behavior provide mixed results. Christensen, et al. (2016) provide a
comprehensive review of DIT studies within the accounting field, looking at the
relationship between DIT-2 P scores and several variables, such as gender, political
affiliation, length of professional service, and effectiveness of types of ethics
instruction at the undergraduate level. Abdolmohammadi and Sultan (2002) and
Buono, et al. (2012) found that students with higher DIT-2 P scores were less likely
to engage in insider trading. West, et al. (2004) found that DIT scores did not
correlate consistently with cheating in a natural experiment on observed cheating
behavior. The authors suggested an investigation into more results than only DIT
P scores and N2 scores. Similarly, Bailey, et al. (2010) encourage a more wholistic
application of DIT survey output that investigates how respondents use moral
reasoning to make decisions.
Following Bailey, et al.’s (2010) suggestion, we focus on four other
measures calculated from the DIT-2 responses in the current study. Two provide
an indication of how consistent a respondent’s ethical reasoning is when making
Flynn, Deno and Buchan
126
value choices regarding relative importance of the statements accompanying the
scenario, and two relate to how well and how consistently ethical reasoning is
applied to the decision made. The four measures are described in the paragraphs
that follow.
The two variables relating to moral reasoning consistency measure
whether responses display consistent application of a moral level or whether
responses appear to indicate transitional thinking between levels of moral
reasoning (Thoma & Rest, 1999; Derryberry & Thoma, 2005). The Type Indicator
variable defines seven types, three for consolidated thinking for each of the levels
(pre-conventional (Type 1), conventional (Type 4), post-conventional (Type 7)),
and four for transitional thinking that leans more toward one level than the other
(e.g., Type 2 is transitional between pre-conventional and conventional but uses
more pre-conventional than conventional thinking while Type 3 is also transitional
between those two levels but uses more conventional than pre-conventional
thinking, and so on). Finally, the Type Indicator score is also condensed into a
Transitional/Consolidated score that groups respondents into either transitional or
consolidated thinking, regardless of level (i.e. 1 for transitional type indicators and
2 for consolidated type indicators).
The two variables that relate to application of ethical reasoning to
decisions made are the “Can’t Decide” score and the Utilizer score. The DIT-2
provides scores on the number of times respondents choose the option of “Can’t
Decide” when pondering the appropriate action to take in a scenario. As there are
5 scenarios in the DIT-2, the number of can’t decide responses may range from
zero to five. This variable indicates how much impact the ranked statements have
on the respondent’s action choice when deciding what to do in the given scenario.
The Utilizer score is an indication of how well the ranking of statements (and the
level of moral reasoning indicated thereby) corresponds to the decision choice,
assuming the respondent did not reply with a “can’t decide” answer. A high
Utilizer score would indicate that the action (and the level of reasoning represented
by that action choice) corresponds highly to the ranking of important statements in
making the judgment (and the level of reasoning represented by the rankings).
Since the Utilizer score cannot be computed if the respondent chose “Can’t
Decide”, there is no Utilizer score for any respondent with either four or five can’t
decide responses for the five scenarios, and the Utilizer score is 0 for any
respondent with three can’t decide responses out of the five scenarios (Bebeau &
Thoma, 2003).
SAMPLE AND FINDINGS
The sample consisted of 63 usable responses from accounting students in
an intermediate and an upper level accounting class. The overall level of moral
reasoning is provided in the N2 score, with a higher score indicating more post-
conventional thinking. In other words, there is an assumption in Kohlberg’s (1969)
cognitive moral development theory that higher levels of moral reasoning are
preferable. Commonly, respondents score at the conventional level of moral
Journal of Business and Accounting
127
reasoning. In the current sample there were no significant differences in average
N2 scores between the intermediate and upper level students.
The frequency of type for the sample is provided in Table 1 below. The
shading highlights the three stages of moral development by grouping transitional
types with the consolidated type closest to it. A total of 23% of the respondents
reasoned predominantly at the pre-conventional stage, a total of 40% of the
reasoned predominantly at the conventional stage, and nearly as many (37%)
reasoned predominantly at the post-conventional stage of moral reasoning.
Table 1: Grouping of Respondents by Type Indicator Score (Total N =
63)
Type Description N Percent Cumulative %
1 Consolidated, Pre-Conventional 1 1% 1%
2 Transitional, Mainly Pre- Conventional
14 22% 23%
3 Transitional, Mainly Conventional 12 19% 42%
4 Consolidated, Conventional 8 13% 55%
5 Transitional, Mainly Conventional 5 8% 63%
6 Transitional, Mainly Post- Conventional
15 24% 87%
7 Consolidated, Post-Conventional 8 13% 100%
Combining the scores into Consolidated vs. Traditional, a total of 17
respondents showed consolidated thinking and the other 46 respondents
demonstrated transitional thinking, thus 27% and 73% as consolidated and
transitional, respectively. This indicates that the students in the sample are not
consistently using one level of moral reasoning when they rank statements of
importance and when they decide upon the appropriate action for the five
scenarios.
Information about the number of can’t decide choices and the utilizer
scores for the sample is provided in Table 2. Since there are five scenarios in the
DIT-2 instrument, the number of can’t decide responses may range from zero to
five. Utilizer scores for the sample for zero, one, and two can’t decide responses
are shown. The utilizer score for three can’t decide options is zero, and there is no
ability to calculate a utilizer score for those respondents answering that they can’t
decide for four or five of the scenarios.
Flynn, Deno and Buchan
128
Table 2. Number Can’t Decide Answers and Utilizer Scores (Total N =
63)
Number of Can’t Decide Responses
Number of Respondents
U
tilizer
Score
N2 Score
Correlation: Utilizer and N2
0 (5 Decisions Made) 21 .18 29.8 -.353
1 (4 Decisions Made) 15 .14 30.1 -.425
2 (3 Decisions Made) 15 .19 25.5 -.281
3 (2 Decisions Made) 4 0 20.8 ---
4 (1 Decision Made) 3 --- 30.6 ---
5 (0 Decisions Made) 5 --- 34.9 --- Several things are noteworthy from the information provided in Table 2.
Fifty-one out of 63 respondents (81%) made a decision about what to do in a
majority of the scenarios. N2 scores are provided in the table to show the level of
moral reasoning by number of decisions made. It is unexpected to find that for the
respondents with a utilizer score different from zero, there is a negative correlation
between the utilizer score and the N2 score. In other words, respondents who made
a decision that more greatly corresponded to their ranking of importance of
statements about the scenario had lower N2 scores than respondents who did not
utilize their rankings to make a decision about what to do in a given scenario. The
ability to evaluate statements for their importance in helping to decide what to do
in a scenario and then to make a decision that is consistent with that evaluation (as
measured by the Utilizer score) is a desired outcome. To find a negative correlation
between N2 score and Utilizer score is and unexpected and undesired outcome of
the study.
Also of significant note is that although subsample size is very small, the
total of five respondents who made no decisions of an action to be taken in any of
the scenarios had the highest average N2 score of any of the other can’t decide
categories. This is consistent with the finding that higher utilizer scores were
associated with lower N2 scores. The group with the highest number of can’t
decide responses had the highest levels of moral reasoning, and the group that
made only one decision out of the five had the second highest mean N2 score.
Table 3 is a combination of the consistency measures presented separately
in Tables 1 and 2. It provides the number of can’t decide choices by Type Indicator
for the 63 respondents. It is a more detailed investigation of consolidated and
transitional thinking at each moral reasoning level paired with the ability to make
a choice about the most appropriate action to be taken.
The numbers below show a large dispersion of Type Indicator
representation for respondents that made three or more choices (0, 1, or 2 can’t
decide responses). Although much smaller numbers of respondents made less than
three choices (3, 4, or 5 can’t decide responses), those respondents are also
dispersed across Type Indicator categories. These results confirm that there is no
Journal of Business and Accounting
129
definitive relationship between level of moral reasoning categorized by Type and
ability to make a decision of an ethical nature.
Collapsing Type Indicator score into Consolidated vs. Transitional and
then grouping by number of can’t decide choices provides summarized
information regarding consistency of moral reasoning as it pertains to decision-
making ability. That information is presented in Table 4.
Table 3: Number of Can’t Decide Responses Grouped by Type Indicator
Type/No. of Can’t Decide (CD)
CD = 0 out of 5
CD = 1 out of 5
CD = 2 out of 5
CD = 3 out of 5
CD = 4 out of 5
CD = 5 out of 5
Total by Type
1: Pre-conv. Consolidated
1 0 0 0 0 0 1
2: Pre-conv. Transitional
6 2 3 1 1 1 14
3: Conv. Transitional
2 4 4 1 0 1 12
4: Conv. Consolidated
2 3 2 1 0 0 8
5: Conv. Transitional
4 0 1 0 0 0 5
6: Post-conv. Transitional
1 6 4 1 2 1 15
7: Post-conv. Consolidated
5 0 1 0 0 2 8
Total by No. of CD Choices
21 15 15 4 3 5 63
Table 4: Number of Can’t Decide Responses by Consolidated vs.
Transitional Classification
# of CD Choices
0 1 2 3 4 5 Total
Total 21 15 15 4 3 5 63
Consolidated 8 (38%) 3 (20%) 3 (20%) 1 (25%) 0 2 (40%) 17 (27%)
Transitional 13 (62%) 12 (80%) 12 (80%) 3 (75%) 3 (100%) 3 (60%) 46 (73%)
Flynn, Deno and Buchan
130
When Type Indicator is collapsed into only consolidated and transitional
classifications across all numbers of can’t decide choices, respondents are
overwhelmingly transitional in their moral reasoning. It is interesting to note that
at the extreme ends of both making a decision in all five scenarios and making a
decision in none of the five scenarios, the number of respondents in the two
classifications of consolidated vs. transitional comes closest to being distributed
equally. The four other can’t decide categories (from one to four decisions made)
range from a low of 75% to a high of 100% of respondents classified as transitional.
For one final parsing of the information, of the 51 respondents making a decision
a majority of the time (less than three can’t decide responses), 37 (73%) were
classified as transitional; of the 12 respondents not making a decision a majority
of the time, 9 (75%) were classified as transitional. It appears that consolidated
moral thinking has no impact on ability to make a decision.
DISCUSSION AND CONCLUSION
The findings presented here indicate that students do not consistently
apply ethical frameworks when making decisions with moral content in a survey
situation. There was virtually no correlation between the moral statements students
had judged to be important in the scenario and the decision that was made, and
most students did not exhibit a consolidated, consistent application of moral
reasoning level when answering questions of an ethical nature. The study adds to
the literature suggesting the need to investigate other components of Rest’s model
and other aspects of ethical behavior (Buono, et al., 2012; Jeffries & Lu, 2018;
Shawver & Sennetti, 2009) and it also investigates other variables provided by
DIT-2 results beyond the P score and N2 score (Bailey, et al., 2010).
The generally accepted approach to business ethics instruction has been to
provide a structured, multi-step model of analyzing an ethical situation in order to
clearly identify issues, options, probable outcomes of the options, and the best
choice to procure the best outcome. However, a formalized, academic approach
may not equip students with all of the tools needed to properly decipher complex
ethical dilemmas fraught with uncertainty and professional pressures. Our findings
add to the literature suggesting that students may not be graduating from
accounting and business programs with the mature understanding needed to reason
about ethical scenarios and then choose a logical action that flows from the
reasoning applied. The areas of immediate investigation would be whether there is
a breakdown in students’ ability to consistently apply ethical frameworks to
analyze specific situations; and whether there is a disconnect between consistent
reasoning (once achieved) and consistent application of the reasoning used to
determine the final action choice. Further research into these areas will hopefully
provide greater insight and avenues for improved ethics education.
Journal of Business and Accounting
131
REFERENCES
Abdolmohammadi, M.J. and J. Sultan (2002). “Ethical reasoning and the use of
insider information in trading.” Journal of Business Ethics. 37 (2): 165-
173.
Bailey, C.D., I. Scott, and S.J. Thoma (2010). “Revitalizing accounting ethics
research in the neo-Kohlbergian framework: Putting the DIT into
perspective.” Behavioral Research in Accounting. 22 (2): 1-26.
Bebeau, M.J. and S.J. Thoma (2003). Guide for DIT-2 – Center for the study of
ethical development. (Minneapolis: University of Minnesota Press).
Buono, A.F., D. Fletcher-Brown, R. Frederick, G. Hall, and J. Sultan (2012).
“Acting ethically: Moral reasoning and business school student behavior.”
SAM Advanced Management Journal. (Summer): 19-26.
Christensen, A.L., J. Cote, and C.K. Latham (2016). “Insights regarding the
applicability of the defining issues test to advance ethics research with
accounting students: A meta-analytic review.” Journal of Business Ethics.
133: 141-163.
Derryberry, W.P. and S.J. Thoma (2005). “Functional differences: comparing
moral judgement development phases of consolidation and transition.”
Journal of Moral Education. 34 (1): 89-106.
Jeffries, F.L. and Y. Lu (2018). “Emotional intelligence as in influence on ethical
behavior: A preliminary study.” Journal of Behavioral and Applied
Management. 19-32.
Jones, D.A. (2009). “A novel approach to business ethics training: Improving
moral reasoning in just a few weeks.” Journal of Business Ethics. 88: 367-
379.
Kohlberg, L. (1969). “Stages and sequences: the cognitive developmental
approach to socialization.” In D. Goslin (ed.), Handbook of Socialization
Theory and Research. (Chicago: Rand McNally).
Rest, J. (1979). Development in Judging Moral Issues. (Minneapolis: University
of Minnesota Press).
Rest, J. (1986). Moral Development: Advances in Research and Theory. (New
York: Praeger).
Shawver, T.J. and J.T. Sennetti (2009). “Measuring ethical sensitivity and
evaluation.” Journal of Business Ethics. 88: 663-678.
Flynn, Deno and Buchan
132
Thoma, S.J. and J.R. Rest (1999). “The relationship between moral decision
making and patterns of consolidation and transition in moral judgment
development.” Developmental Psychology. 35 (2): 323-334.
West, T., S.P. Ravenscroft, and C.B. Shrader (2004). “Cheating and moral
judgment in the college classroom: A natural experiment.” Journal of
Business Ethics. 54: 173-183.
Journal of Business and Behavioral Sciences
Vol 12, No 1; Fall 2019
133
FRAUD-DETECTING EFFECTIVENESS OF
MANAGEMENT AND EMPLOYEE RED FLAGS AS
PERCEIVED BY THREE DIFFERENT GROUPS OF
PROFESSIONALS
Glen D. Moyes University of Central Oklahoma
Asokan Anandarajan New Jersey Institute of Technology
Allen G. Arnold University of Central Oklahoma
ABSTRACT
All organizations are vulnerable to the risk of financial statement fraud. Companies
try to mitigate the threat of fraud by establishing preventive measures. These
measures include implementing specific policies, procedures, programs, and
training to identify potential fraud. These measures comprise the first tier line of
defense against financial statement fraud. The second tier line of defense involves
the establishment of red flags that could ostensibly identify the threat of fraud
occurring. However, not all red flags are equally effective in detecting fraud. The
purpose of this study was to explore how effective different types of flags are in
detecting fraud as perceived by managers. Research was conducted by
questionnaires sent to three distinct groups of professional financial statement
users: Chief Executive Officers and Chief Financial Officers; Controllers and
Treasurers; and Human Resource Directors. This represents a broad swathe of
professional users. These three different groups of professional users were surveyed
to obtain a broad range of perceptions regarding the effectiveness of different red
flags to detect fraud. Perceptions of surveyed professionals identified most
common red flag indicators. The findings of this study are useful in indicating, to
all users, which specific red flags are considered the most important by experienced
professional users. Generally, red flags represent warnings that something could be
wrong or hidden from evaluators. All users of financial statements need to be aware
of the most effective red flags in order to monitor a potentially disastrous situation
and then take immediate corrective measures, as needed.
Keywords: Fraud Detection, Management/Employee Red Flags, CEOs/CFOs,
Controllers/Treasurers, HR Directors
Journal of Business and Behavioral Sciences
134
INTRODUCTION
A red flag is a set of circumstances that are unusual in nature or that vary from
normative activity. It is a signal that something out of the ordinary needs to be
investigated further. Dinapoli (2006) notes that red flags do not indicate guilt or
innocence but merely provide possible warning signals of fraud. The American
Institute of Certified Public Accountants (AICPA) Statement of Auditing Standards
(SAS) No. 99 highlighted the critical importance of fraud investigation. This
auditing standard that was adopted by the Private Company Accounting Oversight
Board (PCAOB) requires external auditors, assisted by internal auditors, to use 42
specific red flags to detect fraudulent activities in conducting financial statement
audits of publicly held corporations. This study focuses on another set of 20
management red flags and 13 employee red flags that were established by the New
York State Controller, Thomas P. DiNapoli. This article explores the level of
effectiveness of these management and employee red flags in detecting fraud as
perceived by 203 respondents, which include the three groups: Chief Executive
Officers and Chief Financial Officers, Controllers and Treasurers, and Human
Resource Directors. Being able to recognize red flags is necessary not only for
auditors and accountants but also for any executive business professional working
in the public sector, where the potential for fraud to occur exists.
LITERATURE REVIEW
The effectiveness of red flags to detect fraud may be directly related to the number
of years of both external auditing experience and internal auditing experience,
based on positional roles of executive and upper-middle corporate management.
Bonner and Lewis (1990) confirmed that more auditing experience improves the
effectiveness of auditors, since more experienced auditors have a greater ability to
apply their auditing knowledge than less experienced auditors, as concluded by
Choo and Trotman (1991). Auditing knowledge is acquired from years of auditing
experience, according to Wright and Wright (1997) and Prawitt, Smith, and Wood
(2008). Similarly, Shamki and Alhajri (2017) verified a positive relationship
between internal auditor effectiveness and their number of years of auditing
experience. Moreover, Bond and DePaulo (2008) found highly experienced
auditors demonstrate 67 percent accuracy in making auditing decisions, and 78
percent accuracy in making auditing decisions not involving fraud.
Many red flag studies focus exclusively on managers. Hackenbrack (1993)
discovered that external auditors of large audit clients concentrate more on the
fraud-risk attributes concerning opportunities for perpetuators to commit fraud.
Apostolou and Hassell (1993) found internal auditors perceived the importance of
red flags in identifying the possible occurrence of management fraud. Similarly,
Albrecht and Romney (1986) validated the significance of red flags as predictors of
management fraud as determined by audit partner respondents. This study indicates
Moyes, Anandarajan and Arnold
135
one-third of red flags are considered significant predictors of fraud and most red
flags tend to focus on the personal characteristics of management. Also, Apostolou,
Hassell, Webber, and Sumners (2001) found external and internal auditors rated
higher fraud indicators concerning management characteristics and their influence
over corporate control environment. Furthermore, Heiman-Hoffman and Morgan
(1996) surveyed external auditors of the then "Big Six" international public
accounting firms and they ranked the thirty most important warning signs of
possible fraud, of which management attitudes are considered the most important
red flags. Church, McMillan, and Scheider (2001) noted internal auditors perceive
red flags as more predictive when managers attempt to falsely overstate earnings in
order to earn bonuses.
METHODOLOGY
The study was conducted by means of a survey instrument. Red Flag
Questionnaires were sent to various types of business professionals to determine
the degree of effectiveness of management and employee red flags in detecting
fraud. The respondent professionals were Chief Executive Officers and Chief
Financial Officers, Controllers and Treasurers, and Human Resource Directors.
The red flag questionnaire used a six-point Likert scale to measure the
professionals’ perceptions of the level of effectiveness of red flags in detecting
fraud. Each professional evaluated the fraud-detecting effectiveness of each red
flag by selecting one of the following responses: not effective, slightly effective,
moderately effective, mostly effective, effective, or extremely effective. If the
professional decided the red flag was extremely effective in detecting fraud, the
value of 6 was entered into the database. In contrast, if the professional decided the
red flag was not effective in fraud detection, a value of 1 was entered into the
database.
The red flag questionnaire was completed and returned by 203 professional
respondents. The 203 respondents who completed the questionnaire are classified
into the following three groups: (1) Chief Executive Officers and Chief Financial
Officers, (2) Controllers and Treasurers, and (3) Human Resource Directors.
Breakdown of the respondents are shown in Tables 1 and 2. The red flags used in
the questionnaire were derived from an article entitled, “Red Flags for Fraud.” This
article listed numerous red flags used to detect fraud and was written by Thomas P.
Dinapoli, New York State Comptroller (n.d.).
Journal of Business and Behavioral Sciences
136
Table 1: THREE DISTINCT CATEGORIES OF PROFESSIONAL
RESPONDENTS
Positional Titles of Professional Respondents Number of
Respondents
Percentages of
Respondents
Chief Executive Officers (CEOs) and Chief
Financial Officers (CFOs)
92 45.3%
Controllers and Treasurers 61 30.1%
Human Resource (HR) Directors 50 24.6%
Totals 203 100%
Table 2 shows various professional certifications of the respondents. The Certified
Public Accountants (CPA) certification represented the largest number with 81
(39.9%) respondents, followed by Chartered Financial Analysts (CFA) at 26
(12.8%); Certified Management Accountants at 18 (8.9%); and Certified Internal
Auditors (CIA) at 18 (8.9%) respondents. The rest of the respondents, at 60
(29.5%), had non-accounting/non-finance certifications.
Table 2: QUANTITY AND TYPES OF CERTIFICATIONS HELD BY
RESPONDENTS
Types of Certifications Number of
Certifications
Percentages of
Certifications
Certified Public Accountants (CPAs) 81 39.9%
Chartered Financial Analyst (CFAs) 26 12.8%
Certified Management Accountants (CMAs) 18 8.9%
Certified Internal Auditors (CIAs) 18 8.9%
Non-Accounting/Non-Finance Certifications 60 29.5
Totals 203 100%
The types of fraud detected by the respondents’ organizations were classified into
21 categories, according the 203 respondents in Table 3. The types of fraud that
occurred the most frequent were embezzlement, credit card fraud, and theft. The
next grouping included nine types of fraud that occurred with regular frequency:
internet, identity theft, money laundering, tax fraud, falsified documents, computer,
billing, payroll, and misappropriation of assets. Finally, a group of seven types of
fraud detected at the lowest frequency included: fraudulent transactions, insurance,
check forgery, misappropriation of funds, unauthorized purchases and payments,
unauthorized signatures, and double payments. This does not indicate that the latter
frauds are not significantly occurring; merely that these frauds have not been as
frequently identified by the respondents in our study.
Moyes, Anandarajan and Arnold
137
Table 3: TYPES OF FRAUD DETECTED BY RESPONDENTS
Types of Fraud Detected
by Respondents
Number of Times
Each Fraud Type
Was Detected
Percentage of Each
Fraud Type Detected
Embezzlement 38 20.9%
Credit Card Fraud 27 14.8%
Theft 18 9.9%
Other Fraudulent Activities 17 9.3%
Fraudulent Financial Reporting 16 8.8%
Internet Fraud 10 5.5%
Identity Theft 7 3.9%
Money Laundering 5 2.8%
Tax Fraud 5 2.8%
Fraudulent Documents 5 2.8%
Computer Fraud 5 2.8%
Billing Fraud 5 2.8%
Payroll Fraud 5 2.8%
Misappropriation of Assets 4 2.2%
Fraudulent Transactions 3 1.7%
Insurance Fraud 3 1.7%
Check Forgery 2 1.1%
Misappropriation of Funds 2 1.1%
Unauthorized Purchases and
Payments
2 1.1%
Unauthorized Signatures 2 1.1%
Fraudulent Double Payment 1 0.55%
Totals 182 100%
The 203 respondents evaluated the level or degree of the fraud-detecting
effectiveness of the following 20 management red flags using a six-point Likert
scale. Table 4 shows the perceived level of effectiveness of each of the 20
management red flags in detecting fraud. The level or degree of fraud-detecting
effectiveness of each management red flag was determined by computing the
mathematical average, or mean, of all the response from the 203 respondents. The
20 management red flags are listed in the descending order of fraud-detecting
effectiveness as shown in the right most column of Table 4.
Management red flags represent a higher priority than employee red flags. Since
managers have the authority to over-ride internal controls and employees do not,
detection of fraud committed by management is considered critically more
important to detect than fraud committed by employees. Based on our survey, the
most significant indicators of management fraud are missing or unaccounted for
documents, financial transactions that are not logical and appear not to make sense,
contracts signed without any tangible evidence of output in the form of products,
Journal of Business and Behavioral Sciences
138
reluctance to provide information to auditors, management decisions dominated by
a single individual or a small group, and high turnover of external auditors. For
further detailed analysis, see Table 4
Table 4: FRAUD DETECTING EFFECTIVENESS OF MANAGEMENT RED FLAGS
Management
Red Flags
Degree of Effectiveness of Red Flags in Detecting Fraud
Not
Effective
Slightly
Effective
Moderately
Effective
Mostly
Effective
Effective Extremely
Effective
Mean
Missing
Documents 17 24 32 35 35 60 4.12
Financial
Transactions
Do Not Make
Sense
18 16 45 33 44 47 4.03
Contracts
Resulting In
No Products
17 22 41 40 47 36 3.92
Management
Reluctance to
Provide
Information to
Auditors
18 26 52 29 41 37 3.79
Management
Decisions
Dominated by
Single
Individual or
Small Group
12 32 40 50 45 24 3.77
Frequent
Changes in
Banking
Accounts
20 27 45 35 42 34 3.76
Frequent
Changes in
External
Auditors
16 34 48 39 33 33 3.68
Unexpected
Overdrafts and
Declines in
Cash Balances
19 33 44 38 37 32 3.67
Management
Engaged in
Frequent
Disputes with
Auditors
24 33 36 39 48 23 3.61
Weak Internal
Control
Environment
27 24 49 34 39 30 3.61
Decentralize
Without
Adequate
Monitoring
22 26 51 41 42 21 3.58
Moyes, Anandarajan and Arnold
139
Compensation
Excessively
Too High
24 28 49 41 35 26 3.56
Excessive
Number of
Year End
Transactions
20 29 55 44 33 22 3.53
Accounting
Personnel Are
Lax and
Inexperienced
with Their
Duties
25 30 50 34 39 25 3.53
Management
Refusal to Use
Pre-Numbered
Documents
24 29 52 43 25 30 3.52
Significant
Downsizing in
Healthy
Market
24 31 50 37 37 24 3.51
Company
Assets Sold
Under Market
Value
23 38 44 41 27 30 3.50
Excessive
Number of
Checking
Accounts
26 32 44 43 33 25 3.49
Continuous
Rollover of
Loans
23 30 54 40 33 23 3.49
Management
Displays
Significant
Disrespect for
Regulatory
Body
28 29 46 44 35 21 3.45
Table 5 shows the perceived level of effectiveness of each of 13 employee red flags
in detecting fraud. The level or degree of fraud-detecting effectiveness of each
management red flag was determined by computing the mathematical average, or
mean, of all the responses from the 203 respondents. These 13 employee red flags
are listed in the descending order of fraud-detecting effectiveness as shown in the
right most column of Table 5. In our survey, the top employee-related red flags are
carrying unusually large sums of cash, providing unreasonable responses to
questions, evidence of heavy gambling, creditors or collectors appearing in the
work place, bragging about significant new purchases, refusing promotions, lack of
segregation of duties among others.
Journal of Business and Behavioral Sciences
140
Table 5: FRAUD DETECTING EFFECTIVENESS OF EMPLOYEE RED FLAGS
Employee Red
Flags
Degree of Effectiveness of Red Flags in Detecting Fraud
Not
Effective
Slightly
Effective
Moderately
Effective
Mostly
Effective
Effective
Extremely
Effective
Mean
Carrying
Unusually
Large Cash
14
27
40
36
44
42
3.96
Providing
Unreasonable
Responses to
Questions
15
29
36
45
44
34
3.87
Gambling
Beyond Ability
to Stand the
Loss
23
20
40
41
38
41
3.86
Creditors or
Collectors
Appearing in
Workplace
21
23
40
41
46
32
3.81
Bragging about
Significant New
Purchases
22
18
53
31
47
32
3.78
Refusing
Promotions in
Fear of
Detection
16
28
46
45
36
32
3.75
Lack of
Segregation of
Duties
17
26
49
42
41
28
3.73
Employee
Lifestyle
Changes
18
29
49
41
30
36
3.71
Excessive
Drinking or
Other Personal
Problems
20
29
43
44
37
30
3.68
Easily Annoyed
at Reasonable
Questions
17
31
51
36
42
26
3.66
High Employee
Turnover 20 34 43 41 38 27 3.61
Refusal for
Vacation or
Sick Leave
28
32
51
45
22
25
3.37
Borrowing
Money from
Co-Workers
24
36
43
38
38
24
3.50
Moyes, Anandarajan and Arnold
141
Tables 6 and 7 indicate the level or degree of the effectiveness of each red flag in
detecting fraud as perceived by each of the three main groups of respondents. In
Table 6, the fraud-detecting effectiveness of each of the 20 management red flags
are showed by each of the three groups: Chief Executive Officers and Chief
Financial Officers, Controllers and Treasurers, and Human Resource Directors. In
Table 6, the level or degree of fraud-detecting effectiveness of each management
red flag is determined by: (1) Mathematical average, or mean, of the responses from
92 Chief Executive Officers and Chief Financial Officers, (2) Mathematical
average, or mean, of the responses from 61 Controllers and Treasurers, and (3)
Mathematical average, or mean, of the responses from 50 Human Resource
Directors.
In general, the Chief Executive Officers and Chief Financial Officers group
evaluates 20 management red flags as slightly more effective in detecting fraud than
the Controllers and Treasurers group. Likewise, the Controllers and Treasurers
group evaluated the management red flags as slightly more effective than the
Human Resource Directors group.
Across the board, management involved red flags include financial transactions not
making any sense; missing documents; management decisions dominated by a
single individual or small groups; unexpected overdrafts, frequent changes of bank
accounts, frequent changes in external auditors, higher compensation to
management relative to industry averages; management engaged in frequent
disputes with auditors, among others. For more detail, see Table 6.
Table 6: AVERAGE FRAUD DETECTING EFFECTIVENESS OF EACH
MANAGEMENT RED FLAG
Management Red Flag Indicators
Chief
Executive
Management
Controllers and
Treasurers
Human
Resource
Directors
Financial Transactions Do Not Make
Sense 4.26 3.90 3.76
Missing Documents 4.20 4.12 3.96
Management Decisions Dominated by
Single Individual or Small Group 4.01 3.74 3.33
Unexpected Overdrafts and Declines
in Cash Balances 3.99 3.52 3.24
Contracts Resulting in Not Products 3.96 3.91 3.84
Management Reluctance to Provide
Information to Auditors 3.93 3.74 3.57
Frequent Changes of Bank Accounts 3.92 3.69 3.53
Frequent Changes in External Auditors 3.84 3.62 3.43
Compensation Excessive Too High 3.75 3.50 3.24
Journal of Business and Behavioral Sciences
142
Weak Internal Control Environment 3.74 3.71 3.24
Management Engaged in Frequent
Disputes with Auditors 3.73 3.59 3.39
Decentralization without Adequate
Monitoring 3.73 3.59 3.29
Excessive Number of Year End
Transactions 3.69 3.48 3.27
Management Refusal to Use
Pre-Numbered Documents 3.74 3.47 3.16
Accounting Personnel are Lax and
Inexperienced with their Duties 3.67 3.52 3.27
Company Assets Sold Under Market
Value 3.65 3.40 3.33
Excessive Number of Checking
Accounts 3.64 3.45 3.27
Management Display Significant
Disrespect for Regulatory Agencies 3.59 3.52 3.10
Continuous Rollover of Loans 3.55 3.55 3.29
Significant Downsizing in Healthy
Market 3.55 3.57 3.37
In Table 7, the fraud-detecting effectiveness of each of the 13 employee red flags
are showed by each of the three groups: Chief Executive Officers and Chief
Financial Officers, Controllers and Treasurers, and Human Resource Directors. In
Table 7, the level or degree of fraud-detecting effectiveness of each employee red
flag is determined by: (1) Mathematical average, or mean, of the responses from 92
Chief Executive Officers and Chief Financial Officers, (2) Mathematical average,
or mean, of the responses from 61 Controllers and Treasurers, and (3) Mathematical
average, or mean, of the responses from 50 Human Resource Directors.
In general, the Chief Executive Officers and Chief Financial Officers group
evaluates 13 employee red flags as slightly more effective in detecting fraud than
the Controllers and Treasurers group. Likewise, the Controllers and Treasurers
group evaluated the employees red flags as slightly more effective than the Human
Resource Directors group.
Table 7: AVERAGE FRAUD DETECTING EFFECTIVENESS OF EACH
EMPLOYEE RED FLAG
Employee Red Flag Indicators
Chief
Executive
Management
Controllers
and
Treasurers
Human
Resource
Directors
Employees Carrying Unusual Large
Sums of Cash 4.21 3.95 3.49
Employees Providing Unreasonable
Responses to Questioning 4.15 3.59 3.65
Moyes, Anandarajan and Arnold
143
Creditors and Collectors Appearing in
Workplace 4.00 3.72 3.53
Employees Bragging about Significant
New Purchases 3.96 3.86 3.35
Employees Gambling Beyond Ability to
Sustain the Loss 3.94 3.86 3.69
Lack of Segregation of Duties in More
Vulnerable Areas 3.94 3.76 3.29
Employee Lifestyle Changes 3.93 3.60 3.41
Employees Refusing Promotions in Fear
of Detection 3.89 3.74 3.51
Employees Excessive Drinking and
Other Personal Bad Habits 3.84 3.62 3.45
Employees Easily Annoyed at
Reasonable Questions 3.84 3.59 3.37
High Employee Turnover in More
Vulnerable Areas of Committing Fraud 3.83 3.43 3.39
Employees Refusal to Take Vacation or
Sick Leave 3.76 3.36 2.63
Employees Borrowing Money from Co-
Workers 3.71 3.45 3.16
The most common red flags for fraud committed by employees include employees
providing unreasonable responses to questioning, creditors and collectors appearing
in the work place, lack of appropriate segregation of duties, constantly refusing
promotion, and refusal to take vacation or sick leave among others.
In Table 8, the 203 respondents reported 27 different ways that fraud was detected
by their organizations. As shown in Table 8, the most common ways fraud was
detected was by informants and internal auditors. The most common methods were
informants followed by internal auditors, fraudulent documents, missing funds,
auditing of online transaction activities, among others. For more details, see Table
8.
Table 8: DIFFERENT WAYS THAT FRAUD WAS DETECTED BY
ORGANIZATIONS
Different Ways Fraud Was Detected Number Percentage
Informants 50 14.8%
Internal Auditing 38 11.2%
Fraudulent Transactions and Fraudulent
Documents 32 9.5%
Missing Funds 25 7.5%
Random Sample of Documents and Files 21 6.3%
Audited Online Transactional Activities 14 4.2%
Security System Checks and Analysis 14 4.2%
Personal Purchases Using Company Funds 14 4.2%
Journal of Business and Behavioral Sciences
144
Unauthorized Purchases and Unauthorized
Payments 13 3.9%
Computer System Checks and Analysis 12 3.6%
Unauthorized Credit Card Charges 11 3.3%
Mathematical Discrepancies and Not Balancing 10 3.0%
Fraudulent Purchases and Altered Documents 9 2.7%
Missing Documents 9 2.7%
Compared Credit Card Receipts and Bank
Statements 9 2.7%
Audited Timecards and Electronic Surveillance 7 2.1%
Tax Discrepancies 6 1.8%
Auditing Software Analysis 6 1.8%
Electronic Analysis of Data 6 1.8%
Inventory Shortages 6 1.8%
Fraudulent Contractors 5 1.5%
Billing and Invoice Discrepancies 4 1.2%
IT Department Detection 4 1.2%
Fraudulent Wire Transfers 3 .9%
Perpetuators Arrested and Prosecuted 3 .9%
Fraudulent Computer Logging In Access
Attempts 2 .6%
Monthly Closing Processes 2 .6%
Totals 335 100%
In Table 9, the 203 respondents disclose 26 different types of evidence that proves
fraud was committed.
Table 9: TYPES OF EVIDENCE THAT PROVED FRAUD WAS
COMMITTED IN ORGANIZATIONS
Types of Evidence Proving Fraud Was Committed Number Percentage
Missing Funds 75 19.4% Fraudulent Documents 44 11.4% Informants 34 8.8% Fraudulent Credit Card Receipts 27 7.0% Fraudulent Transactions 21 5.5% Computer System Information Output 18 4.7% Internal Auditing Evidence 18 4.7% Fraudulent Computer or Online Transactions 18 4.7% Missing Documents 12 3.1% Fraudulent Vendor Invoices 12 3.1% Fraudulent Canceled Checks 11 2.9% Security System Information Output 11 2.9% Computer System Information Output 10 2.6%
Moyes, Anandarajan and Arnold
145
Not Balancing 10 2.6% Fraudulent Payroll Documents 9 2.3% Fraudulent Tax Documents 8 2.1% Bank Documents, Statements and Reconciliations 7 1.8% Missing Inventory Items 6 1.6% Unauthorized Purchase Transactions 5 1.3% Unauthorized Signatures on Documents 5 1.3% Auditing Software Evidence 5 1.3% Spreadsheet Analysis of Data 5 1.3% Police Detected/Arrested and District Attorneys
Prosecuted 4 1.0%
Missing Assets 4 1.0% IT Department Detected Fraud 3 .8% Fraudulent Wire Transfer Documents 3 .8% Totals 385 100%
CONCLUSIONS
The purpose of this study was to examine which financial red flags are the most
effective in detecting fraud as seen from the eyes of three groups of professional
financial statement users: Chief Executive Officers and Chief Financial Officers,
Controllers and Treasurers, and Human Resource Directors. The findings of this
study has substantive validity in that this study’s respondents, Chief Executive
Officers and Chief Financial Officers, Controllers and Treasurers, and Human
Resource Directors, are highly qualified and experienced financial statement users
with stellar professional certifications. Their views based on survey responses
roughly converged highlighting the importance of the red flags addressed in this
research.
The types of fraud detected by the organizations were classified into 21 categories
according to the 203 respondents in Table 3. The type that occurred the most was
embezzlement, credit card fraud, and theft. According to Chief Executive Officers
and Chief Financial Officers, Controllers and Treasurers, and Human Resource
Directors, the most frequent management-involved red flags, in Table 4, include
missing documents, financial transactions not making any sense, contracts resulting
in no products, management reluctance to provide information to auditors, and
management decisions dominated by a single individual or small groups. In Table
5, the top four employee related red flags are carrying unusually large sums of cash,
providing unreasonable responses to questions, evidence of heavy gambling, and
creditors or collectors appearing in the work place. Employee behaviors send
warning signals to vigilant internal auditors and interested parties. These signals
could include lavish changes in life style, increased propensity for drinking and
gambling, and borrowing money from co-workers, among others.
Journal of Business and Behavioral Sciences
146
The top three indicators of average fraud detecting effectiveness of red flags were
financial transactions not making any sense, missing documents, management
decisions dominated by a single individual or small groups for management red
flags, as shown in Table 6. Employee red flag average fraud detecting
effectiveness, in Table 7, were employees carrying unusual large sums of cash,
employees providing unreasonable responses to questioning, and creditors and
collectors appearing in workplace. Table 8 displayed the top four ways that fraud
was detected by organizations: informants, internal auditing, fraudulent
transactions and fraudulent documents, and missing funds. Table 9 focused on the
types of evidence proving that fraud was committed and the four prevalent
categories were missing funds, fraudulent documents, informants, and fraudulent
credit card receipts.
This research is obviously limited by the relatively small sample size (n = 203), but
the quality of the respondents indicates a very high level of reliability, as delineated
by their professional titular roles and credentials, as well as their earned
certifications. This study’s findings should lead to further research into usefulness
of indicating which specific red flags are considered the most important by
experienced professional users. All users of financial statements need to be aware
of the most effective red flags in order to monitor a situation and then take
immediate corrective measures, as needed.
REFERENCES
Albrecht, W. S., and Romney, M. B. (1986). Re-flagging management fraud: A
validation. Advances in Accounting, 3, 323-333.
Apostolou, B., and Hassell, J. M. (1993). An empirical examination of the
sensitivity of the analytical hierarchy process to departures from
recommended consistency ratios. Mathematical and Computer Modeling,
17(4-5), 163-170.
Apostolou, B., and Hassell, J. M., Webber, S. A., and Sumners, G. E. (2001). The
relative importance of management fraud risk factors. Behavioral
Research in Accounting, 13, 1-24.
Bond, C. F., and DePaulo, B. M. (2008). Individual differences in judging
deception: Accuracy and bias. Psychological Bulletin, 134, 477-492.
Bonner, E., and Lewis, B. L. (1990). Determinants of auditor expertise. Journal of
Accounting Research, 28, 1-20.
Moyes, Anandarajan and Arnold
147
Choo, F., and Trotman, K. T. (1991). The relationship between knowledge
structure and judgments for experienced and inexperienced auditors. The
Accounting Review, 66(3), 464-485.
Church, B. K., McMillan, J. J. and Schneider, A. (2001). Factors affecting internal
auditors’ consideration of fraudulent financial reporting during analytical
procedures. Auditing: A Journal of Practice and Theory, 20(1), 65-80.
Dinapoli, T. P. (n.d.). Red Flags for Fraud. Retrieved from
https://www.osc.state.ny.us/localgov/pubs/red_flags_fraud.pdf
Hackenbrack, K. (1993). The effect of experience with different sized clients on
auditor evaluations of fraudulent financial reporting indicators. Auditing:
A Journal of Practice and Theory (Spring), 99-110.
Heiman-Hoffman, V. and Morgan, K. P. (1996). The warning signs of fraudulent
financial reporting. Journal of Accountancy (October), 75-77.
Prawitt, D. F., Smith, J. L. and Wood D.A. (2008). Internal audit quality and
earnings management. Institute of Internal Auditors Research
Foundation, 1-42.
Shamki, D. and Alhajri T.A. (2017). Factors influence internal audit function.
International Journal of Business and Management, 12(10), 143-154.
Wright, A. and Wright S. (1997). The effect of industry experience on hypothesis
generation and audit planning decisions. Behavioral Research in
Accounting, 9, 1-28.